1 THE IMPACT OF PLANT DIVERSITY ON ARTHROPOD COMMUNITIES WITHIN VEGETABLE PRODUCTION SYSTEMS IN FLORIDA AND HAITI By HEIDI NO L HANSPETERSEN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 Heidi No l HansPetersen
3 To Jeff ak moun lakay!
4 ACKNOWLEDGMENTS I would like to thank my husband, Jeff who followed my dreams al l the way to Haiti. I am grateful for my Gainesville family : the Campbells, Mosers, MacMasters, Stigge Kaufmans, Pat Galiger, Brennemans Selkes, Dixons, and others from Emman uel Mennonite Church, who dependably provided fun bike rides, excellent meals, engaging conversation, and a community in which I have truly felt beloved. My friends in Haiti welcomed Jeff and me into thei r lives taught us their language, music, and insects T his warm welcome changed us for the better. I am especially thankful for my Chair Dr. McSorley and coworkers Rosie, Romy, Namgay, Simon, and John Fredrick who were a source of encouragement to me and made coming to work a pleas ure. I appreciate the guidance and insight offered by my committee members: Dr. Carlene Chase, Dr. Pete Hildebrand, and Dr. Oscar Liburd. Debbie Hall has helped me with so many details over the past five years, and I am in her debt. Lyle Buss assi sted me with several insect identifications. Finally I wo uld like to acknowledge my parents who nurtured my curiosity and made me spend a lot of time outdoors! My dad taught me my first insect less ons with a fly rod in hand, and piqued my interest in b iology while collecting caddis fly larvae. My mom continues to inspire me with her vocational passion which I hope to model in my future work.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 8 ABSTRACT ................................................................................................................... 10 CHAPTER 1 INTRODUCTION .................................................................................................... 13 The Need for Pesticide Alternatives ........................................................................ 13 Vegetational Diversity and Arthropod Populations .................................................. 15 Weeds as a Source of Vegetational Diversity ......................................................... 16 Research Objectives and Hypotheses .................................................................... 19 2 THE IMPACT OF WEEDING REGIMES ON ARTHROPOD COMMUNITIES IN PEPPER, Capsicum annuum ................................................................................. 21 Introduction ............................................................................................................. 21 Materials and Methods ............................................................................................ 23 Field Management and Experimental Design ................................................... 23 Arthropod Sampling .......................................................................................... 24 Yields ............................................................................................................... 26 Weeds .............................................................................................................. 26 Statistical Analysis ............................................................................................ 27 Results .................................................................................................................... 28 Weeds .............................................................................................................. 28 Arthropod Sampling .......................................................................................... 29 Yields ............................................................................................................... 32 Discussion .............................................................................................................. 32 3 THE IMPACT OF INTERC ROPPING SQUASH WITH NONCROP VEGETATION BORDERS ON THE ABOVE GROUND ARTHROPOD COMMUNITY .......................................................................................................... 48 Introduction ............................................................................................................. 48 Materials and Methods ............................................................................................ 49 Field Management and Experimental Design ................................................... 49 Arthropod Sampling .......................................................................................... 51 Statistical Analysis ............................................................................................ 53 Results .................................................................................................................... 53 Border Composition .......................................................................................... 5 3 Arthropod Sampling .......................................................................................... 54 Discussion .............................................................................................................. 58
6 4 PEST MANAGEMENT STRATEGIES AND LIVELIHOOD ACTIVITIES OF SMALL SCALE AGRICULTURALISTS IN HAITIS NORTHERN CENTRAL PLATEAU ............................................................................................................... 77 Introduction ............................................................................................................. 77 Rationale ................................................................................................................. 78 Methodology ........................................................................................................... 79 Results .................................................................................................................... 81 Study Area ........................................................................................................ 81 Climate and Environment ................................................................................. 82 Cropping System .............................................................................................. 83 Agricultural Landscape ..................................................................................... 85 Training ............................................................................................................ 92 Challenges to Agricultur al Production .............................................................. 93 Pest management ...................................................................................... 93 Seed conservation ..................................................................................... 98 Disease management .............................................................................. 100 Soil amelioration and plant nutrition ......................................................... 102 Household Values and G oals ......................................................................... 103 Recommendations ................................................................................................ 105 Community Associations ................................................................................ 106 Focus on the Lakou ........................................................................................ 107 Extend the Harvest ......................................................................................... 108 Erosion Mitigation ........................................................................................... 108 Integrated Pest Management Training ........................................................... 109 Target Future Farmers ................................................................................... 112 Conclusions .......................................................................................................... 112 5 THE IMPACT OF FOUR WEED MANAGEMENT STRATEGIES ON INVERTEBRATE POPULATIONS IN CENTRAL HAITI ........................................ 121 Introduction ........................................................................................................... 121 Materials and Methods .......................................................................................... 123 Invertebrate Sampling .................................................................................... 124 Statistical Analysis .......................................................................................... 125 Results .................................................................................................................. 125 Discussion ............................................................................................................ 126 6 SURVEY OF INVERTEBRATE FAUNA OF SOIL CONSERVATION BARRIERS IN CENTRAL HAITI .............................................................................................. 132 Introduction ........................................................................................................... 132 Materials and Methods .......................................................................................... 133 Study Area ...................................................................................................... 133 Invertebrate Sampling .................................................................................... 134 Results .................................................................................................................. 135 Discussion ............................................................................................................ 135
7 7 CONCLUSIONS ................................................................................................... 142 LIST OF REFERENCES ............................................................................................. 147 BIOGRAPHICAL SKETCH .......................................................................................... 159
8 LIST OF TABLES Table page 2 1 Weed composition of treated plots (0.5 m2 sample), 200708. ........................... 38 2 2 Number of taxa (richness) collected in pitfall traps and pan traps during 2007 and 2008 growing seasons. ................................................................................ 38 2 3 Effect of weed treatments on selected arthropod groups (number per trap) recovered in pitfall traps, 2007. ........................................................................... 39 2 4 Effect of weed treatments on selected arthropod groups (number per trap) recovered in pitfall traps, 2008. ........................................................................... 40 2 5 Orthogonal contrasts between unweeded control and 100% weeded treatment for selected arthropod groups captured (number per trap) by various sampling methods, 2007, 2008. ............................................................. 41 2 6 Effect of treatments on selected arthropod groups (number per trap) collected in pan traps, 2007. ............................................................................... 42 2 7 Effect of treatments on selected arthropod groups (number per trap) collected in pan traps, 2008. ............................................................................... 43 2 8 Capture of selected arthropod groups (number per trap) on yellow sticky cards in weed treatments, 2007. ......................................................................... 44 2 9 Capture of selected arthropod groups (number per trap) on yellow stick y cards in weed treatments, 2008. ......................................................................... 45 2 10 Effect of treatment on selected arthropod groups (number per 8 plants) observed during in situ counts, 2008. ................................................................. 46 2 11 Effect of weed treatments on pepper yields (number fruit per plot), 2007. ......... 47 2 12 Effect of weed treatments on pepper yields (number fruit per plot), 2008. ......... 47 3 1 Border treatment mean height and plant stand one week prior to transplanting and one month after transplanting. ................................................ 64 3 2 Selected beneficial arthropod groups (number per trap) recovered on sticky cards, pan traps and pitfall traps within border treatments, 2007. ...................... 65 3 3 Selected arthropod groups (number per trap) recovered on sticky cards, pan traps and pitfall traps within border treatments, 2007. ........................................ 66 3 4 Selected beneficial arthropod groups (number per trap) recovered on sticky cards, pan traps and pitfall traps within border treatments, 2008. ...................... 68
9 3 5 Selected arthropod groups (number per trap) recovered on sticky cards, pan traps and pitfall traps w ithin border treatments, 2008. ........................................ 69 3 6 Selected beneficial arthropod groups (number per trap) recovered on sticky cards, pan traps and pitfall traps within squash crop, 2007. ............................... 71 3 7 Selected arthropod groups (number per trap) recovered on sticky cards, pan traps and pitfall traps recovered in adjacent squash crop, 2007. ........................ 72 3 8 Selected beneficial arthropod groups (number per trap) recovered on sticky cards, pan traps and pitfall traps within squash crop, 2008. ............................... 74 3 9 Effect of treatments on selected arthropod groups (number per trap) recovered on sticky cards, pan traps and pitfall traps recovered in adjacent squash crop, 2008. ............................................................................................. 75 4 1 Fruit and nut trees found in greater Bohoc region. ........................................... 115 4 2 Field crops cultivated in rainfed areas of the greater Bohoc region. ................. 116 4 3 Vegetables reported by informants to be cultivated in the study area. ............. 117 4 4 Pest management techniques as reported by informants. ................................ 118 4 5 Agricultural pests descr ibed by informants. ...................................................... 119 4 6 Insecticides used in the greater Bohoc region. ................................................. 120 5 1 The influence of four weed management tactics invertebrate numbers per pitfall trap. ......................................................................................................... 130 5 2 The influence of four weed management tactics on invertebrate numbers per board trap. ........................................................................................................ 131 6 1 Survey site information. .................................................................................... 139 6 2 Invertebrate numbers per pitfall trap placed in living erosion control borders ( ranp vivan). ...................................................................................................... 140 6 3 Invertebrate numbers per pitfall trap placed in rock erosion control borders ( misek) ............................................................................................................. 141
10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE IMPACT OF PLANT DIVERSITY ON ARTHROPOD COMMUNITIES WITHIN VEGETABLE PRO DUCTION SYSTEMS IN FLORIDA AND HAITI By Heidi No l HansPetersen May 2010 Chair: R. McSorley Major: Entomology and Nematology Reliance on agricultural chemicals as the primary method of pest control is not a practical longterm pest management strategy, especially for low resource farmers in developing countries like Haiti where poverty, illiteracy, and limited access to training compound the risks associated with pesticide use. Conservation biological control is an ecologically based pest management strategy that seeks to increase exis ting beneficial arthropods and enhance their regulatory activity by making the agricultural lan dscape more favorable for their establishment and survival In this study, the impact of vegetational diversity on the abundance of pest and natural enemy populations was studied in vegetable production systems in Florida, and central Haiti. Noncrop vegetation including weeds and refuge strips was incorporated into squash ( Cucurbita pepo L.), pepper ( Capsicum annuum L.), and mixed cropping systems. Additionally the pest management practices of farmers living in rural Haiti were investigated using the S ondeo method. In Florida, pest and natural enemy abundance, weed richness and biomass, and yields were compared in pepper in Florida with different weed densities using multiple sampling techniques. Beneficial insects were generally found in greater num bers in the
11 unweeded treatment than in weedfree plots. Greater numbers of arthropod taxa correlated with increased weed richness. However, leaving weeds within the plot did not result in increased biological control and negatively a ffected pepper yields A similar experiment conducted in Haiti compared the influence of weed density and locally available mulches on the invertebrate community. The use of sugarcane ( Saccharum sp.) bagasse mulch was more useful than maintaining a weed refuge for increasing numbers of beneficial arthropods in the soil surface community and reduced the labor associated with weeding. The impact of intercropping noncrop vegetation on the aboveground arthropod community was assessed in squash. Natural enemies were most abundant in the native weed complex and pigeon pea borders; however there was only limited spillover of natural enemies into the adjacent squash crop. Intercropped strips did not influence the movement of thrips and whiteflies into the plots. None of the border intercrop treatments prevent ed a heavy infestation of melonworm ( Diaphania hyalinata L ) Conversational interviews conducted during a participatory rural appraisal in north central Haiti revealed that agrochemicals ranging from moderat ely to highly toxic are available to rural Haitian farmers in the study area witho ut label and safety information. Currently, pesticides are only used for seed conservation and in highvalue vegetable production. Farmers in the study area have received l ittle training in pest management and none of the farmers interviewed were aware of beneficial insects. Integrated pest m anagement (IPM) training would be highly beneficial to farmers in this community. Successful training is likely to take place within existing community groups and among future farmer groups.
12 A survey of selected invertebrate groups associated with different soil conservation barriers in central Haiti was conducted during f all 2009. Pitfall traps were used to characterize the inverteb rate communities within contour hedgerows and nonliving erosion barriers in both successional and mixed cropping systems. Collembola, ants and spiders were most closely associated with the barrier habitat regardless of the type of border or cropping sys tem. Local farmers that participated in the survey expressed interest in the ecosystem services provided by natural enemies and fauna involved in decomposition and nutrient cycling.
13 CHAPTER 1 INTRODUCTION The Need f or Pesticide Alternatives The effect of a single pesticide application is rarely confined to its target organism. The disruptive action of pesticides disequilibriates the complex food webs that sustain both natural and agricultural ecosystems and may impose risks to human health. Reliance on agricultural chemicals as the primary method of pest control is not a practical longterm pest management strategy especially for low resource farmers in developing countries like Haiti where poverty, illiteracy, and limited access to training compound the risks associated with pesticide use. Despite recommendations by FAO and WHO, highly toxic agrochemicals are available to peasant farmers in countries that lack the political infrastructure to safely regulate pesticide use (Goldmann et al. 2004). Thus, some of the most hazardous agrochemicals are being used by those least equipped to cope with the environmental and health costs associated with their use. As a consequence, 99% of acute pesticide poisonings occur in developing nations (Goldmann et al. 2004). In addition to the accumulation of toxic residues in the environment by some pesticides and the development of resistant pests, the disruptive action of broadspectr um pesticides leads to imbalance in the agroecosystem (Lewis et al.1997). Broadspectrum pesticides kill the beneficial organisms that perform key ecosystem functions including the regulation of pests, decomposition, and pollination (Lewis et al. 1997, Al tieri 2002). Without the stabilizing presence of natural enemies, pest populations often surge to higher levels than prior to the chemical application (Lewis et al. 1997, Flint and Gouveia 2001). The ecological effects of chemical based pest
14 management str ategies create a built in need for more pesticide use (Murray 1994). O n ce farmers start using broadspectrum agricultural chemicals, they will often have to increase the dosage or number of sprays to achieve similar results to the fi rst application. This phenomenon, described as the pesticide treadmill ( Van Den Bosch 1978), results in increased risk of exposure to both the applicator and the environment. Integrated pest management (IPM) is an ecosystem based strategy that combines multiple tactics to prevent unacceptable levels of pest damage ( Geier 1966). Rather t han relying on pesticides alone, IPM tactics focus on the modification of farming practices to decrease pest problems physical control measures that disrupt pest environments (cultural controls), and the use of beneficial organisms to regulate pests (biologic al control) (Luckmann and Metcalf 1994) Together with careful monitoring, these methods can significantly reduce and sometimes eliminate the need for chemical control measures The benefits derived from these strategies include a decrease in environmental and health risks associated with pesticide exposure ( Luckmann and Metcalf 1994). T he elimination or reduction of broadspectrum pesticide applications and the establishment of favorable habitat s within the agroecosystem can promote the survival, reproduc tion dispersal and ultimately the regulatory activities of natural enemies (Landis et al. 2000, Flint and Gouveia 2001 ). However, their success depends in part on the timely availability of food resources, adequate shelter and alternate prey or hosts. The development of this ecological infrastructure is the basis of habitat management, a form of conservation biological control that seeks to preserve and enhance existing natural enemy populations by altering the agricultural landscape (Landis et al. 2000). Since many predators and parasitoids rely on alternate prey and
15 plant provided nutrition like pollen, nectari es and plant sap (Wilkens o n and Landis 2000), habitat management often involves increasing vegetational diversity. Diversification can occur spatially, temporally, and across multiple layers within the agroecosystem (Landis et al. 2000). On a landscape level this may involve the conservation of nearby noncrop habitats ( Tscharnkte et al. 2007), while at the farm level, hedgerows, adjacent fields and boundaries can be managed to enhance diversity. Within a crop, diversification can be increased through various intercropping strategies or manipulating weed populations (Altieri and Letourneau 1982). As such, polyculture is a salient fea ture of traditional cropping systems (Altieri 2002). Vegetational Diversity and Arthropod P opulations Conventional agricultural systems in most developed countries are typically characterized by uniform plantings of a single crop species. For specialist herbivor es, sole cropping provides a concentrated supply of food, shelter, oviposition sites, and potential mates (Root 1973, Altieri and Letourneau 1982) that together create optimal conditions for a pest outbreak (Altieri and Nichols 2004, Altieri and Letourneau 1982, Pimenta l 1961). Both spatially and temporally diversified crops, more common in smallholder agriculture in developing countries reduce the concentration of these requisites such that the agricultural landscape becomes less favorable to invasion (R oot 1973, Altieri and Letourneau 1982). Vegetational diversity increases the complexity of the agroecosystem in term s of plant architecture, habitat connectivity, and odor profiles Together, these influence the abundance, distribution, and activity of p ests and natural enemies alike ( Nyoike and Liburd 2010, Altieri 1984, Norris and Kogan 2000, Capinera 2005). Additional complexity may also encourage the exploitation of microhabitats not available in
16 monocultures and enhance alternate food and prey avail ability (Andow 1991, Sheehan 1986). The potential mechanisms behind these phenomena have been extensively reviewed (Hooks and Johnson 2003, Smith and McSorley 2000, Andow 1991, Russell 1989, Sheehan 1986). In a review of empirical literature pertaining to the impact of polyculture on arthropods, Andow (1991) concluded that more than half of the herbivorous species studied decrease in polycultures, while over 30% of species studied responded variably or did not change. This trend was reversed for more than 50% of the natural enemy species whose populations were more abundant in polycultural systems, while only 9.3% were found in lower densities (Andow 1991). Although the diversification of agricultural systems does not uniformly result in fewer pests and greater natural enemy activity, vegetational diversity is widely believed to contribute to greater ecosystem stability (Altieri and Nichols 2004 a ) and minimizes risks associated with pest outbreaks (Pimental 1961). Weeds as a Source o f Vegetational Diversity The preeminence of chemical control methods within conventional agriculture promotes the expectation of complete weed removal from the syst em (Maxwell and ODonovan 2007). Failure to remove weeds is often viewed as a poor cultural practice (Hillocks 1998). Many f armers are reluctant to intentionally leave weeds in fields, field margins, or fence rows because of their potential to reduce yields, produce weed seed, contaminate harvested products, and harbor insects or plant diseases that may be d etrimental to crops (Maxwell and ODonovan 2007). Low resource and organic farmers have fewer weed control options available, and therefore, may be forced to
17 tolerate some level of weed infestation. Ineffective weed control is the key obstacle for grower s transitioning from conventional to organic farming (Brberi 2002). Hand weeding is the most common management technique in developing countries (Hobbs and Bellinder 2004). Failure to completely remove weeds is often seen as t he result of labor constraints. H owever, in many traditional agricultural systems, certain weeds are deliberately left in association with crops, despite the potential yield losses that might occur (Altieri et al. 1987, Hillocks 1998). In a typical Haitian kitchen garden it is n ot uncommon to see pigweed, A maranthus spp. being allowed to seed, so that new tender leaves can be harvested and eaten as a vegetable. In fact, the Haitian Kreyl word for Amarathus spp. translates as spinach and includes both domesticated amaranth and wild species. Subsistence farmers who rely on weeds as a source of animal forage, botanical pharmaceuticals and for home consumption often practice nonclean cultivation (Altieri et al. 1987, Hillocks 1998). This relaxed weeding regime (Altieri et al. 1987) provides weedy refugia within a vegetable cropping system that may impact invertebrate activity. Despite problems associated with weeds, their presence is not always deleterious t o crop yields (Bugg et al. 1987, Andow 1988 Schellhorn and S ork 1997). Futhermore, weeds provide several ecosystem services that have high value, especially within the context of agricultural systems with low external inputs such as organic and subsistence farming (Gliessman 1988). Weeds rapidly colonize disturbed areas providing a protective soil covering that mitigates erosion, increases water infiltration, slows the leaching of nutrients, and contribut es to soil organic matter (Gliessman 1988). Under some conditions, retaining weeds within crop production syst ems has had a net positive
18 effect on crop production due to their impact on pest populations and natural enemy activity (Andow 1988, Kemp and Barrett 1989, Wolcott 1928). Unlike most pest species, weeds are primary producers within the agricultural ecosyst em (Marshall et al. 2002) that provide foliage, roots, shoots, fruits, nectar, pollen, weed seeds, and refuge for arthropods across various trophic levels. These plant provided resources serve as a significant nutritional energy source for predators and parasitoids that can enhance their establishment, reproduction, dispersal, and survival during periods of prey scarcity (Landis et al. 2005, Wilkenson and Landis 2005, Altieri and Whitcomb 1980) as well as their regulatory activity in both annual and perennial cropping systems (Altieri and Letourneau 1982; Andow 1988, 1991). One of the simplest ways to increase diversity and plant resources within a cultivated field is by allowing the presence of some weeds (Landis et al. 2005). Vegetational diversity pr ovided by weeds has been associated with greater abundance and diversity of predacious arthropods and parasitoids (Letourneau 1987; Andow 1991; Showler and Greenburg 2003). Incorporating weedy refuge strips within agricultural fields has been a successful measure for increasing beneficial arthropods (Showler and Greenburg 2003, Netwig 1998), and for enhancing their activity such that yields are impacted positively ( Showler and Greenburg 2003, Landis et al. 2000, Kemp and Barrett 1989). Furthermore, certai n pest outbreaks are more likely to occur in weedfree fields than in diversified agricultural systems that retain noncrop weeds (Altieri and Whitcomb 1980; Andow 1991; Bezerra et al. 2004). Both ecological and economic consequences are associated with a llowing weeds to remain in a crop. Their presence both spatially and temporally within the agricultural ecosystem plays an important role in the management of insects (Norris and Kogan
19 2000, 2005). W hen weeds are retained within the agricultural landscape, they often are responsible for yield losses that outweigh any benefits associated with pest control (Andow 1991) Therefore, farmers should carefully consider the impact of weedcrop competition on harvest quality and the potential suite of pests and pl ant pathogens that may be associated with endemic weeds. Research Objectives and Hypotheses Because conservation biological control is one of the most accessible IPM tools for low resource farmers (Altieri 2002), experiments were designed with crops and habitat manipulations that could be easily integrated into existing cropping systems in rural Haiti including handweeding, locally available mulches, and intercropping. Experiments were first conducted for two growing seasons in Florida (Chapters 2 and 3) and then adapted for a third season in Haiti (Chapter 5) In Chapter 2, w e examined the impact of retaining different proportions of the natural weed complex on the aboveground arthropod community. By using various sampling techniques, we hoped to understand the impact of four different weeding regimes on multiple arthropod families, including pests and natural enemies within the agroecosystem. Our hypothesis was that retaining different weed densities would effect the aboveground arthropod community and yield of pepper ( C. annuum ). Futhermore, weedy treatments would have greater numbers and ri chness of arthropods than weedfree treatments. This experiment was modified and carried out in Haiti and discussed in Chapter 5. In Chapter 3, we examined the influence of four border treatments on the aboveground arthropod community. Our hypotheses were that border crop composition would a ffect populations of arthropods in an adjacent crop by functioning as a barrier to the
20 movement of weak flying insects, and serving as colonization sites for pests and natural enemies that spill over into the crop. Chemicalbased pest management and the use of highly toxic pesticides have been promoted in the southern vegetable producing region of Haiti (Bishop 1995). Ecological damage rendered by these pesticides was also a salient feature of the agricultural practices of the neighboring Dominican Republic (Murray 1994). In 1987 and 1988, the Dominican Republic had the highest rate of illegal pesticide (i.e. DDT) residues on foods imported into the US (Murray 1994). Because of the porous nature of the border betw een the Dominican Republic and Haiti, I chose to investigate whether obsolete chemicals had reached the communities surrounding Bohoc in north central Haiti as well as the pest management techniques being employed by rural farmers in this region (Chapter 4) Since erosion is the greatest source of land degradation in Haiti (Lewis and Coffey 1985), organizations throughout the country have been promoting the construction of soil conservation barriers ( Bayard et al. 2007). Permanent erosion control barrier s offer uncultivated habitat within the agricultural landscape that can provide resources and refuge for the invertebrate community. A survey of erosion control barr iers was conducted to determine if they served as benefical arthropod refugia (Chapter 6).
21 CHAPTER 2 THE IMPACT OF WEEDING REGIMES ON ARTHROPOD COMMUNITIES IN PEPPER, Capsicum annuum Introduction One of the simplest ways to increase diversity and plant resources within a cultivated field is by allowing the presence of some weeds (Landis et al. 2005). Unlike most pest species, weeds are primary producers within the agricultural ecosystem (Marshall et al. 2002) that provide foliage, roots, shoots, fruits, nectar, pollen, weed seeds, and refuge for arthropods across various trophic levels. Vegetational diversity provided by weeds has been associated with greater abundance and diversity of predacious arthropods and parasitoids (Letourneau 1987, Andow 1991, Showler and Greenburg 2003). Integrating weeds into the agricultural landscape provides several of the requisites for the establishment, reproduction, and dispersal of natural enemies (Altieri and Whitcomb 1980) that can enhance their regulatory activity in both annual and perennial cropping systems (Altieri and Letourneau 1982, A ndow 1988, 1991). Weeds contribute additional complexity in terms of plant architecture, habitat connectivity, and odor profiles that influence the abundance, distribution, and activity of pests and natural enemies alike. They influence host finding by i nterfering with both visual and olfactory cues and can delay colonization of a crop by some insect pests (Altieri 1984, Norris and Kogan 2000; Capinera 2005). Furthermore, certain pest outbreaks are more likely to occur in weedfree fields than in diversi fied agricultural systems that retain noncrop weeds (Altieri and Whitcomb 1980, Andow 1991, Bezerra et al. 2004). Herbivores that have the ability to exploit resources from both crop and weed species may benefit from the increased diversity that weeds c ontribute to agricultural
22 ecosystems (Andow 1988). However, this is not always detrimental to crops. When weeds were left in a tomato field, infestation levels of Bemesia tabaci were lower compared with tomatoes planted alone (Bezerra et al. 2004). Lower pest infestation levels in weedy fields may be the combined effect of host preference (Wolcott 1928, Showler and Greenburg 2003, Capinera 2005), the effects of natural enemies, and a dilution effect (Bezerra et al. 2004). Within the context of an integr ated approach to pest management, it is essential for growers and researchers to consider the impact that weed removal may have on arthropod population dynamics (Schellhorn and Sork 1997). Careful monitoring of insect activity on weeds and appropriate tim ing of weed management practices are necessary to effectively control insect pests in diverse systems (Capinera 2005, Norris and Kogan 2000, 2005). Both ecological and economic consequences are associated with allowing weeds to remain in a crop. When weeds are retained within the agricultural landscape they often are responsible for yield losses that outweigh any benefits associated with pest control (Andow 1991). However, under some conditions, retaining weeds within crop production systems has had a net positive effect on crop production due to their impact on pest populations and natural enemy activity (Altieri and Whitcomb 1980, Letourneau 1987, Andow 1988, 1991, Kemp and Barrett 1989). Incorporating weedy refuge strips within agricultural fields h as been a successful measure for increasing beneficial arthropods (Showler and Greenburg 2003, Nentwig 1998) and for enhancing their activity (Altieri and Whitcomb 1980, Kemp and Barrett 1989, Landis et al. 2005). Increased plant and arthropod diversity within the agricultural landscape may effectively r educe pest associated losses, contribute to greater community stability, and minimize risk associated with pest outbreaks (Pimental 1961).
23 In the current study, we examine the influence of retaining diff erent proportions of the natural weed complex on the aboveground arthropod community in a diverse agricultural system with similarities to the growing conditions of many low input farmers. By using various sampling techniques, we hoped to understand the impact of four different weeding regimes on multiple arthropod families, including pests and natural enemies within the agroecosystem. The goal of this study was to determine whether retaining different weed densities within a diverse vegetable production system would influence the aboveground arthropod community and yield of pepper ( C. annuum ). Materials and Methods Field Management and Experimental Design Field experiments were conducted at the University of Florida Plant Science Research and Education Unit (PS REU) (29o24N, 82o9W) located in Citra, FL, during the fall growing seasons in 2007 and 2008. The soil was classified as Arredondo sand (95% sand, 2% silt, 3% clay) (Thomas et al. 1979) with 1.5% organic matter. Prior to establishing the exper iment and during the off season, fields were maintained in weedy fallow. In May of 2007, a selective herbicide, Fusilade (Syngenta Crop Protection, Inc., Greensboro, NC) was applied over the experimental area to control annual and perennial grasses In June of 2008, RoundUp ( Monsanto Company, St. Louis, MO) was applied to the experimental area prior to discing. Throughout both growing seasons, an undisturbed weedy reservoir (20 m x 16 m) was preserved at the west ern end of the field to provide continuity and overwintering habitat for arthropods (Drinkwater et al. 1995). Because of the labor associated with hand weeding, i ndividual plots measured 2.4 m x 2.4 m T wo double rows of Aristotle peppers with 1.2 m spacing between rows
24 were transplant ed on 4 September 2007 and 5 August 2008 by hand. They were planted 0.3 m apart in rows resulting in a density of 28 pepper plants per plot. Four weeding regime treatments were replicated four times in a randomized complete block design. The treatments included: unweeded control, 50% of the plot area weeded, 50% of the plot area weeded with the addition of a bean intercrop, and 100% of plot weeded. In both 50% weeded treatments, an area 0.6 m x 2.4 m surrounding each row of peppers was kept weed free. In the remaining area, the natural weed complex was permitted to grow (50% weeded) or beans were planted (bean intercrop). For the intercrop treatment, a row of bush beans, Phaseolus vulgaris (Roma II in 2007 and Blue Lake in 2008), was handseeded on both sides of the weedfree area one week prior to transplanting peppers. All weeding was done manually and the time required to maintain each plot recorded. Peppers were fertigated by drip irrigation with soluble nitrogen at a rate of 16.8 kg N/ ha per week and sidedressed with 2 g granular slow release fertilizer (206 9 Osmocote, N P2 O 5K2 O, Scotts Miracle Gro Company, Marysville, OH) as needed. Twoweeks prior to transplanting peppers, 3 m wide strips of cowpea, Vigna unguiculata, Iron & Clay mixed cowpeas, were drilled mechanically at 56 kg/ha between blocks. Space between plots within each block was manually cultivated with a hoe and by hand to maintain bare fallow. Arthropod Sampling Arthropod communities were sampled using four sampling t echniques: pitfall traps, sticky card traps, pan traps, and in situ counts. All traps were placed randomly in the interior rows of peppers approximately 0.3 m from pepper plants and at least 0.6 m from the edge of the plot between 800 and 1200 EST. Traps were left in the field for 24 h.
25 Pitfall trap s Plastic sandwich containers (14 cm x 14 cm x 4 cm; 532 mL) were used as pitfall traps (Triplehorn and Johnson 2005) and buried so that the upper edge was flush with the soil surface. The traps were filled three quarters (ca. 300 mL) with water, along with 2 mL of dish detergent (Ultra Joy, Procter & Gamble, Cincinnati, OH) to break the surface tension and prevent arthropods from escaping. One pitfall trap was placed in each plot on 2 October, 23 October, and 13 November during 2007 and on 4 September, 23 September, 16 October, and 5 November during the 2008 growing season. Pan traps Clear polyethylene deli containers (11 cm in diameter x 4.5 cm deep; 236 mL) (Gainesville Paper Company, Gainesville, FL) were used as pan traps. A single pan was placed in each plot at midplant height and filled half way with water (ca. 175 mL) along with 2 mL dish detergent. Pan traps were collected from each plot on 19 September, 10 October, 15 October, 29 October, and 13 November during 2007 and on 26 August, 4 September, 24 September, 7 October, and 21 October during 2008. Sticky cards. A single unbaited Pherocon AM trap (Yellow Sticky, YS; Great Lakes IPM, Vestaburg, MI) was placed at plant height in one of the interior pepper rows in each plot on the following dates: 10 October, 25 October, and 5 November 2007; and 25 August, 15 September, 6 October, and 28 October 2008. Collected sticky cards were wrapped in plastic food wrap (Stretchtite, Polyvinyl Films Inc., Sutton, MA) and stored at 4 Insects trapped within the grid (23 cm x 18 cm) on the sticky card were counted and recorded. A representative sample of the whiteflies and thrips on each card was counted using an exclusionary grid that allows one to count 15 of the 63 squares (2.5 cm x 2.5cm) on each card (Finn 2003).
26 In situ counts. Eight plants were selected systematically from each plot for whole plant visual counts. Arthropods were identified to order and family when possible. Visual counts were conducted on 26 September, 11 October, 26 October, 8 Nov ember 2007, and on 15 September, 1 October, 13 October, 20 October and 3 November during 2008. Yields On 17 November 2007, all pepper fruits and graded into the following categories: U.S. Fancy grade (size minimum of 7.62 cm in diameter and at least 8.89 cm long), U.S. No. 1 (minimum diameter and length 6.35 cm), U.S. No.2 Grade (size smaller than U.S. No. 1) (USDA 1989). The temperature dropped to 1.407 which ended the experiment. In 2008, p eppers were harvested, weighed, and graded on 14 October, 21 October, 28 October, 4 November, 10 November, and 18 November. In 2008, culled fruits were categorized as having insect damage or other damage (most often sun damage). Data for culled fruits we re presented as a percentage of the total harvest. Total marketable yield weight included all graded fruits without insect or other damage. The 2008 field season was concluded on 19 November when the temperature dipped to 0.25 Weeds At the end of eac h season, the weed composition of each plot was evaluated by sampling weeds from two randomly placed 0.5m2 quadrats. Weeds were identified to genus level and species when possible, cut at ground level, and dried at 70 days. Weeds were categorized as broadleaf weeds, grasses or sedges, and aboveground weed biomass was calculated for each plot. Weed richness was determined by counting the number of weed species per 0.5 m2.
27 Statistical Analysis Arthropods collected in traps were identified to order and at the family level. In many cases, orders represented by only a few individuals in several families were grouped together by order or feeding guild for statistical analysis. Commonly occurring arthropod groups and one particularly frequent genus were analyzed. Data for all members of the entire arthropod community were reported for each of the sampling methods However, pitfall trapping targets organisms found at the soil surface, while pan trap and stickycards are typically used to sample fl ying insects (Southwood and Henderson 2000) Data from arthropod counts were analyzed by repeated measures analysis of variance (ANOVA) using the GLM procedure (SAS Institute 2008). Means were separated with the least significant difference (LSD) test at P 2008). Preplanned c omparisons between the unweeded control and 100% weeded plots were further explored by orthogonal contrasts ( P LM Pitfall trap data for Formicidae were log transformed by log10 (x+1) prior to analysis to stabilize variance, but untransformed means are reported. Weed composition and weed richness data were subjected to ANOVA using the SAS GLM procedure. Me ans were separated with the LSD at P 0.10 (SAS Institute 2008). Arthropod richness was calculated for pitfall traps and pan traps by counting the number of families caught in each trap, with the exception of the families of microhymenoptera and microdiptera, which were not identified to family. Richness data were analyzed by repeated m easures ANOVA using the SAS GLM procedure and means were separated with the LSD test (SAS Institute 2008). Insect richness data
28 from the final sampling date for pitfall traps and pan traps were correlated with weed richness data collected at the same time, by calculating Pearson Correlation Coefficients ( r ) using SAS (SAS Institute 2008). Results Weeds The native weed complex was allowed to grow naturally in all treatments containing weeds; therefore, the weed composition of each plot was not the same and varied between years. The native weed complex was composed of a mixture of broadleaf weeds, grasses and sedges ( Table 21). During 2007, the native weed complex was more than 60% grasses, primarily crabgrass, Digitaria spp. and goosegrass, Eleusine ind ica (L.) Gaertn, with some barnyard grass, Echinochloa crus galli (L.) Beauv b ermudagrass Cynodon dactylon (L.) Pers., and crowfootgrass, Dactyloctenium aegyptium (L.) Willd. Broadleaf weeds, primarily Florida pusley, Richardia scabra L., followed by hairy indigo, Indigofera hirsuta L., and purslane, Portulaca oleracea L., dominated the native weed complex during the 2008 growing season. The proportions of each of these species did not differ significantly among treatments (Table 21). During 2007, no significant differences were observed in the amount of time required to weed among the 100% weeded, 50% weeded, and bean intercrop treatments (data not shown). However, in 2008, the bean intercrop treatment (2.93 0.24 min./plot) required significantly less weeding time ( F = 45.75; d f = 3, 147; P < 0.0001) than both the 50% weeded (3.76 0.36 min./plot) and the 100% weeded (4.33 0.45 min./plot) treatments. The time required to weed 50% of a plot did not differ
29 significantly from that needed to rem ove 100% of the weeds The three weeded treatments all required more labor than the control. Arthropod Sampling Pitfall traps. No differences in richness occurred among the four treatments in pitfall traps during in during either year ( Table 22). The major families of insects recorded in treated plots included: Aphididae (aphids) Dolicopodidae (longlegged flies) Formicidae (ants) herbivorous Heteroptera (true bugs) Lepidoptera, macrohymenoptera (wasps and bees greater than 1 cm in length), Orthopt era (grasshoppers and crickets) and predatory Coleoptera (Tables 2 3 and 2 4). The group A uchenorrhyncha was comprised primarily of adult and immature members of the family Cicadellidae (leafhoppers) along with fewer numbers within Cercopidae (spittlebugs) Delphacidae (planthoppers) and Membracidae (treehoppers) In bot h 2007 ( P = 0.0089) and 2008 ( P = 0.0184) auchenorrhyncans were most numerous, in the treatments containing weeds and lowest in the 100% weeded treatment (Tables 2 3 and 24 ). During 2008, microdiptera and the collective Diptera group were most abundant in the treatments that contained the most weeds (unweeded and 50% weeded) ( Table 24). This trend was not significant during the previous season. When the 100% weeded and unweeded tr eatments were contrasted, microhym enoptera were more abundant ( P = 0. 0359) in unweeded plots during both 2007 and 2008 ( Table 25). Numbers of predatory Cole optera and Aranae ( spiders ) were higher in the unweeded control than in the 100% weeded treatment in 2008, but not in 2007, when numbers were either low (Coleoptera) or highly variable (spiders). Likewise, total numbers of predators and total numbers of natural enemies were greater
30 in the unwe eded control than in the 100% weeded treatment in 2008 but not in 2007 ( Table 25). A correlation between weed richness and arthropod richness in pitfall traps ( r = 0.599, P d.f. = 14) was observed only in 2008. However, the percentage of Poaceae ( grasses) among weed biomass per plot was associated with pitfall trap arthropod richness in both years ( r = 0.427 in 2007 and r = 0.433 in 2008; P d.f.= 14). In 2008, arthropod richness in pitfall traps was strongly correlated with the percentage of hairy indigo among broadleaf weed biomass ( r = 0.695, P 0.01, d.f = 14). Pan traps. Capture rates in pan traps were fairly low when compared with the numbers of arthropods found in pitfall traps and on sticky cards and significant differences among treatments were detected primarily during the 2008 season (Tables 2 6 and 2 7). In 2008, numbers of dolichopodids, microhymenoptera, and Thysanoptera (thrips) were greatest ( P = 0.0167) in the unweeded control and least in the 100% weeded plots. When the unweeded control and 100% weeded treatments were contrasted (removing the variability from the intermediate treatments), significantly more Auchenorrhyncha, herbivorous Heteroptera, and Dolichopodidae were trapped in the control plots during both the 2007 and 2008 growing seasons ( Table 25). Although Lepidoptera and macrohymenoptera showed some effects from the treatments, fewer than 0.25 arthropods were captured per trap. Arthropod richness among the treatments was higher ( P ) in the unweeded control than in the 100% weeded treatment in both 2007 and 2008 ( Table 22). Weed richness and arthropod richness in pan traps were correlated during both 2007 ( r = 0.492, P < 0.1, d.f. = 13) and 2008 ( r = 0.429, P < 0.10, d.f. = 14). In 2007, insect richness
31 in pan traps was associated with goosegrass biomass ( r = 0.810, P < 0.001, d.f.= 13). Arthropod richness in pan traps was also correlated with the proportion of Cyperaceae (sedges) (r = 0.436; P < 0.10, d.f.= 14) and biomass of Poaceae ( r = 0.433, P < 0.10, d.f.= 14) within the plot during 2007. Sticky cards. During 2007, the only significant response to weeding regime detected on yellow sticky cards occurred with Thysanoptera, primarily Frankliniella spp. Thrips populations were highest ( P = 0.0385) in the intermediate treatments ( Table 28). In 2008, the unweeded control and 50% weeded plots contained more ( P = 0.0009) auchenorrhynchans than the 100% weeded plots ( Table 29). F lea beetles, Altic a sp p. (Chysomelidae: Coleoptera) were easily identified to genus and more abundant in the 100% weeded treatment than the bean intercrop and unweeded control in 2008. When the 100% weeded treatment was contrasted with the unweeded control, more ( P = 0.0187) flea beetles were captured when plots were weedfree during both seasons ( Table 25). The predatory C oleoptera group (Coccinellidae, Cantharidae, Staphylinidae, Mordellidae, and Carabidae) was also more abundant in the 100% weeded treatment when contrasted with the control during both 2007 and 2008 ( Table 2 5). In situ counts. Among all the arthropod groups observed, no differences were observed in situ during 2007 (data not shown). However, when direct counts were conducted in 2008, Altica spp. were most numerous ( P = 0.0307) in the 100% weeded treatment ( Table 210) and dipterans were most abundant in the intermediate treatments.
32 Yields In 2007, no peppers were graded as U. S. Fancy grade fruits and U.S. No. 1 pepper yields did not differ with weeding regime (Table 211) The number of U.S. No. 2 fruit was significantly higher ( P = 0.0083) in the 100% weeded treatment than in the 50% weeded treatment and the unweeded control ( Table 211). In 2008, the number of Fancy and U.S. No. 1 fruits harvested were greatest in the 100% weeded and least in the unweeded treatments ( Table 212). Numbers harvested from the bean intercrop treatment did not differ from those harvested in the 100% weeded plots. In 2008, total marketable yields by weight were highest ( P = 0.0014) in the 100% weeded treatment (5.23 0.92 kg/plot) and bean intercrop (4.8 0.95 kg/plot) followed by the 50% weeded plot (3.19 0.45 kg/plot), and least in the unweeded control (1.17 0.34 kg/plot). Damage to fruit was not detected in Fancy grade fruit, and t he percentage of damaged fruit harvested did not differ among treatments for the other fruit grades (data not shown). Discussion Bell pepper competes poorly with weeds (Lanini and LeStrange 1994), therefore it is not surprising that the unweeded control yi elded the least number of peppers in both seasons. The greatest number of marketable peppers were harvested in the 100% weeded plots, however during 2007 and 2008, both the number of fruits and total marketable weight did not differ from the bean intercr op treatment. Despite few differences in weed cover and arthropod population between the two intermediate treatments, the bean intercrop often outyielded the 50% weeded treatment. It is possible that higher yields associated with this treatment were the result of additional nitrogen provided through the bean intercrop or a reduction in weed competition at the
33 beginning of the season. This effect was brief, as the bean intercrop succumbed to early infestation by whiteflies and subsequent infection by a Begomovirus, which was indicated by severe stunting, chlorosis, mosaic patterning on leaves and confirmed by inclusion body analysis of leaf tissue according to procedures outlined by Christie and Edwardson (1994). Although the beans never reached more than 30 cm in height, they interfered with weeds sufficiently to reduce both yield loss and labor required for weeding in 2008. The beans may have functioned as a short lived living mulch, which is an effective means for reducing weed interference during the critical weed free period (Chase and Mbya 2008), while permitting weeds to provide ecosystem services to the crop later in the season after the crop is established. In 2007, the bean crop succumbed two weeks earlier than in 2008, which was not sufficient to prevent losses from yield or labor costs. Multiple sampling methods were used to obtain information about several members of the aboveground arthropod community. The use of yellow sticky cards is a standard sampling method to determine relative abundance, and is particularly attractive to whiteflies and aphids (Southwood and Henderson 2000). However, colored traps do not provide information regarding the natural movement patterns of insects (Powell et al. 1996). Passive traps including clear pan traps and pitfall traps likely provide more information regarding natural patterns of locomotion (Powell et al. 1996, Southwood and Henderson 2000). Although sticky cards demonstrated that there were no differences among the four treatments in terms of relat ive abundance of microhymenoptera, their movement and activity within the canopy were better described by the results of passive trapping. Significantly more microhymenopterans were detected in the unweeded control than the 100% weeded treatment using pas sive
34 trapping methods. These results suggest that either the additional plant architecture and heterogeneity associated with weeds or the disturbance of cultivation influenced parasitoid movement and behavior within the canopy (Marino and Landis 1996, Landis and Menalled 1998). Because parasitism data were not collected during these experiments, we cannot make any conclusions regarding parasitoid efficiency. However, their natural movement patterns reflected in pitfall and pan traps suggest greater activ ity in pepper when weeds are abundant and in close proximity. It is important to note that pitfall trapping is most commonly used for sampling grounddwelling species. This method may not sample all species with the same efficiency, and the presence of v egetation around the traps can influence capture rates as well (Powell et. al 1996). To further investigate this question of microhymenoptera movement patterns, interception traps such as window traps or transparent sticky cards may be used to monitor the movements including the direction of movement of flying insects (Powell et al. 1996). Although there were rarely differences within the arthropod community between the two intermediate treatments, the 100% weeded treatment and the unweeded control differ ed significantly for several groups. At the ground level, which was monitored with pitfall traps, total natural enemies including microhymenoptera, predatory Coleoptera, and spiders, were generally found in greater numbers in the unweeded control than in the 100% weeded treatment. Auchenorrhycans were consistently associated with the plots that contained the most weeds and were lowest in the 100% weeded plots. Kemp and Barrett (1989) found that more cicadellids occurred in soybeans inter cropped with successional weeds than when intercropped with sown grassy strips. Non predatory Coleoptera, ants, and aphids were unaffected by the weeding regimes with the
35 exception of herbivorous flea beetles, Altica spp. (Chrysomelidae), which occurred in greatest numbers in the 100% weeded treatment. Cu ltivation may encourage outbreaks of particular insects including flea beetles, whose host locating behavior appears to be adapted to bare soil (Cromartie 1975). Flea beetle populations increased in both conventional tillage systems (Dosdall et al. 1999) and in monocultures where vegetational diversity is low (Tahvanainen and Root 1972, Root 1973, Kareiva 1985). Tahvanianen and Root (1972) demonstrated that intercropping interfered with host finding and f eeding behavior of Phyllotreta cruciferae. The habitat simplicity that results from cultivation and the absence of a diverse background may signal favorable conditions to colonizing insects like flea beetles (Root 1973, Cromartie 1975, Dosdall et al. 1999). This may be the case with our research since Altica spp. was often most abundant in the weedfree treatment. Increasing vegetational diversity may have important effects on the density, distribution and diversity arthropods in agricultural systems (N orris and Kogan 2005). Calculating richness is a convenient and simple method for estimating an important component of diversity (Powers and McSorley 2000). We estimated arthropod richness in passive traps only, and observed that richness of taxa using pan traps was greatest in the treatments with the most weeds and lowest in the weedfree treatment. These results suggest that weed richness influences the number of different arthropods moving through the canopy level at the height of the pepper plant. This idea is further supported by the correlation between arthropod richness in pan traps and weed richness. In a discussing the benefits of polyculture, Andow (1991) suggests that the marginal benefits in pest control can be overshadowed by yield losses, more so than in
36 associated monocultures. In this study, competition from weeds was the most limiting factor affecting yield loss. Although insect pests were present, key pepper pests, including pepper weevil ( Anthonomus eugenii ) and caterpillars (Mossler et al. 2006), were rarely observed. Solanaceous weeds that serve as reproductive hosts of A. eugenii (Schuster 2006 ) and disease reservoirs were virtually absent from the experimental area as well. Because the removal of solanaceous weeds from agricult ural fields is a standard recommendation for pest and disease management in pepper (Mossler et al. 2006, Schuster 2006), our results may have been very different had more solanaceous species been present within the agricultural landscape. In cleanweeded fields, habitat complexity is sacrificed to reduce competition and increase yields (Altieri and Whitcomb 1980). Whereas weeddiversified systems provide additional complexity that has often been demonstrated to increase natural enemy populati ons, the weeds may also result in unacceptable yield losses for growers (Altieri and Whitcomb 1980, Bugg et al. 1987, Andow 1988, 1991). We observed that retaining weeds within the experimental plots increased the richness and numbers of several arthropod groups; however, there was no demonstrated impact on biological control. Additionally, weed competition resulted in yield losses. Intermediate treatments, especially the bean intercrop treatment, did not reduce yields; but neither did they demonstrate c lear effects on the arthropod community. By surveying many members of the arthropod community through multiple sampling methods, we observed that the density, presence or absence of weeds within the system does not have the same effect on all arthropod gr oups and that certain groups remained completely unaffected even when the treatment extremes were contrasted. The impact of weeds on key pest organisms is an important consideration for manipulating the agricultural
37 ecosystem through weed management. How ever, weedcrop competition during the early stages of pepper growth impacted yield losses more dramatically than pests in this study. Therefore, retaining weeds for the purpose of supporting natural enemies within the agroecosystem should be done careful ly, to ensure that weeds are not competing with the crop. In this study, a short lived mulch from a bean intercrop along with 50% weeding provided similar levels of natural enemies as the unweeded control yet did not compromise yields.
38 Table 21. Weed c omposition of treated plots (0.5 m2 sample), 200708. Year Treatment % Broadl eaf % Grass % Sedge Biomass ( kg/0.5 m 2) 2007 Control 33.33 13.53 61.16 12.13 5.52 2.67 0.10 0.01 Intercrop 21.16 17.22 76.81 17.96 2.03 0.79 0.09 0.03 50% 31.15 20.67 67.96 20.69 0.89 0.22 0.09 0.01 2008 Control 65.88 14.07 29.71 13.62 4.40 5.02 0.09 0.02 I ntercrop 53.49 17.54 46.52 17.54 0.00 0.00 0.06 0.01 50% 69.38 18.86 27.06 17.96 3.56 4.45 0.06 0.01 Data are means standard errors of 4 replications. No significant differences ( P 0.10) among treatments. Table 22. Number of taxa (richness) collected in pitfall traps and pan traps during 2007 and 2008 growing seasons. Pitfall t raps 1 Pan traps 2 Treatment 2007 2008 2007 2008 C ontrol 10.17 0.72 9.50 0.68 4.48 0.36 a 5.42 0.42 a Bean intercrop 11.75 0.66 10.69 0.78 4.05 0.37 ab 3.71 0.33 ab 50% weeded 11.41 0.75 10.94 0.67 4.40 0.34 ab 4.75 0.35 a 100% weeded 10.75 0.69 9.69 0.81 3.25 0.33 b 3.41 0.31 b A NOVA 3 results F 1.1 0.92 2.5 6.66 P 0.3621 0.4385 0.0644 0.0004 Contrast 4 : 100% weeded vs unweeded control F 0.37 0.03 5.99 15.51 P 0.5438 0.8587 0.0168 0.0002 Data are means standard errors of 4 replications. Means in columns followed by the same letter do not differ ( P 0.10) according to LSD test. No letters in column indicate no differences among means. 1 Primarily species from the orders Diptera, Hempitera and Hym enoptera. 2 Primarily Aranae, Coleoptera, Diptera, Hemiptera and Hymenoptera. 3 Analysis of v ariance; F and P values ; d. f. = 3, 9. 4 d. f. = 1, 9
39 Table 23. Effect of weed treatments on selected arthropod groups (number per trap) recovered in pitfall traps, 2007. Treatment ANOVA 1 Arthropod g roup C ontrol Bean i ntercrop 50% weeded 100% w eeded F value P Dermaptera 1.67 0.90 0.58 0.26 0.17 0.17 1.42 0.74 0.88 0.4604 Orthoptera 2.67 0.54 3.42 0.66 3.75 0.89 3.00 0.63 0.50 0.6800 Hemiptera Aphididae 1.33 0.75 1.08 0.66 0.67 0.33 0.58 0.15 0.45 0.7179 Auchenorrhyncha 1 9.75 1.67 a 11.17 1.7 a 10.75 2.5 a 4.17 0.86 b 4.40 0.0089 Herbiv. Heteroptera 2 0.75 0.28 0.50 0.29 0.58 0.29 0.00 0.00 1.68 0.1871 Coleoptera Non predatory 3 0.58 0.34 1.58 0.51 1.70 0.44 1.58 0.40 1.34 0.2800 Predatory 4 1.08 0.40 1.42 0.62 1.08 0.29 0.33 0.19 1.19 0.3300 Hymenoptera Formicidae 1.38 0.07 1.27 0.93 1.44 0.12 1.30 0.39 1.61 0.4700 Microhymenoptera 5.25 0.87 a 4.92 0.65 b 3.33 0.63 ab 3.08 0.65 b 2.29 0.0900 Macrohymenoptera 0.33 0.22 0.25 0.18 0.33 0.14 0.75 0.25 1.17 0.3300 Diptera Dolichopodidae 2.75 0.91 2.42 0.71 2.67 0.97 1.75 0.37 0.34 0.7961 Microdiptera 6.67 1.63 5.50 1.18 6.33 1.54 6.25 1.59 0.11 0.9600 Total D iptera 9.50 2.22 8.33 1.80 9.25 2.36 8.75 1.88 0.06 0.9789 Araneae 6.00 4.05 7.25 3.42 5.83 3.08 1.67 0.41 0.70 0.5600 Total predators 5 9.75 4.58 11.17 3.60 8.33 3.08 5.00 1.07 0.71 0.5536 Total natural enemies 6 15.33 4.62 16.33 3.56 12.50 3.14 8.83 1.57 1.08 0.3700 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P 0.10) according to LSD test. No letters in row indicate no differences among means. 1 A nalysis of variance; F and P values ; d. f. = 3, 41. 2 Cercopidae, Cicadellidae, Delphacidae, and Membracidae. 3 Primarily Cyndidae, Lygaeidae and Miridae. 4 Chrysomelidae, Elateridae, Scarabaeidae 5 Carabidae, Cantharidae, Coccinellidae, Mordellidae, and Staphylinidae. 6 Predatory Coleoptera, Diptera ( Asilidae, Syrph idae), Hemiptera (Anthrocoridae, Geocorus spp. and Reduviidae), and Dermaptura. 7 Total predators m acro hymenoptera, and micro hymenoptera.
40 Table 24. Effect of weed treatments on selected arthropod groups (number per trap) recovered in pitfall traps, 2008. Treatment ANOVA 1 Arthropod g roup C ontrol Bean i ntercrop 50% weeded 100% w eeded F value P Dermaptera 0.21 0.10 b 1.25 0.31 a 0.94 0.35 ab 1.13 0.29 a 2.95 0.0400 Orthoptera 1.13 0.24 1.13 0.39 0.88 0.24 0.69 0.20 0.63 0.6002 Hemiptera Aphididae 5.81 2.40 1.56 0.55 3.81 2.53 0.38 0.20 1.85 0.1476 Auchenorrhyncha 1 5.46 1.29 a 2.75 0.62 bc 5.31 1.20 ab 2.13 0.45 c 3.62 0.0184 Herbiv. Heteroptera 2 0.77 0.23 0.56 0.24 0.81 0.49 0.19 0.10 0.90 0.4488 Coleoptera Non predatory 3 0.25 0.14 b 1.00 0.26 a 0.38 0.13 b 0.38 0.20 b 3.17 0.0312 Predatory 4 3.25 0.89 2.75 0.70 2.19 0.54 1.50 0.30 1.39 0.2537 Hymenoptera Formicidae 1.21 0.07 0.99 0.07 1.11 0.07 1.09 0.10 1.36 0.2643 Microhymenoptera 4.54 0.74 3.06 0.63 3.56 0.69 2.69 0.37 1.73 0.1703 Macrohymenoptera 0.48 0.16 0.50 0.18 0.13 0.09 0.25 0.14 1.59 0.2016 Diptera Dolichopodidae 8.21 3.37 4.50 1.56 7.06 2.00 2.75 0.77 1.26 0.2957 Microdiptera 5.90 0.85 a 5.50 0.73 a 5.50 0.69 a 2.88 0.40 b 4.33 0.0081 Total Diptera 14.42 3.71 a 10.13 1.90 ab 12.94 2.18 a 5.69 0.76 b 2.49 0.0692 Araneae 3.81 1.80 1.75 0.99 1.69 0.42 0.56 0.18 1.61 0.1976 Total predators 5 7.73 1.93 5.94 1.54 5.06 1.14 3.25 0.67 1.70 0.1773 Total natural enemies 6 12.75 2.05 a 9.50 1.90 ab 8.75 1.28 ab 6.19 0.86 b 2.82 0.0471 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P 0.10) according to LSD test. No letters in row indicate no differences among means. 1 Analysis of variance; F and P values; d. f. = 3, 57. 2 Cercopidae, Cicadellidae, Delphacidae, and Membracidae. 3 Primarily Cyndidae, Lygaeidae and Miridae. 4 Chrysomelidae, Elateridae, Scarabaeidae 5 Carabidae, Cantharidae, Coccinellidae, Mordellidae, and Staphylinidae. 6 Predatory Coleoptera, Diptera ( Asilidae, Syrphidae), Hemiptera (Anthrocoridae, Geocorus spp. and Reduviidae), and Dermaptura. 7 Total predators m acro hymenoptera, and micro hymenoptera.
41 Table 25. Orthogonal contrasts between u nweeded control and 100% weeded treatment for selected arthropod groups captured (number per trap) by various sampling methods, 2007, 2008. 2007 ANOVA 1 2008 ANOVA Arthropod group Unweeded control 100% weeded F -value P Unweeded control 100% weeded F -value P Pitfall traps Microhy menoptera 5.25 0.87 3.08 0.65 4.47 0.0407 4.54 0.74 2.69 0.37 4.62 0.0359 Predatory C oleoptera 1.08 0.40 0.33 0.19 2.53 0.1194 3.25 0.89 1.50 0.30 3.77 0.0570 Araneae 6.00 4.05 1.67 0.41 1.11 0.2991 3.81 1.80 0.56 0.18 4.63 0.0357 Total predators 9.75 4.58 5.00 1.07 1.11 0.2991 7.73 1.93 3.25 0.67 4.91 0.0307 Total natural enemies 15.33 4.62 8.83 1.57 2.02 0.1627 12.75 2.05 6.19 0.86 8.30 0.0056 Pan Traps Dolichopodidae 1.35 0.44 0.50 0.22 3.63 0.0606 1.50 0.34 0.00 0.00 12.6 0.0006 Auchenorrhyncha 1.50 0.44 0.65 0.21 2.90 0.0900 0.58 0.16 0.25 0.11 3.50 0.0647 Herbiv. Heteroptera 1.60 0.69 0.50 0.30 3.78 0.0559 0.33 0.13 0.08 0.06 3.92 0.0507 Sticky card Altica spp. 0.00 0.00 1.00 0.51 5.64 0.0223 0.06 0.06 0.50 0.18 5.86 0.0187 Predatory C oleoptera 0.17 0.11 1.00 0.35 3.28 0.0775 1.25 0.45 2.75 0.86 2.94 0.0921 Data are means standard errors of 4 replications. 1 A nalysis of variance; F and P values ; d. f. = 1, 41 for pitfall traps (2007) and d. f. = 1, 57 (2008); d. f. = 1, 73 for pan traps (2007 and 2008); d. f. = 1, 41 for sticky cards (2007 and 2008).
42 Table 26. Effect of treatments on selected arthropod groups (number per trap) collected in p an traps 2007. Treatment ANOVA 1 Arthropod g roup C ontrol Bean i ntercrop 50% weeded 100% w eeded F value P Coleoptera 0.05 0.05 0.10 0.07 0.10 0.07 0.15 0.08 0.35 0.7873 Diptera Dolichopodidae 1.35 0.44 0.95 0.31 0.80 0.22 0.50 0.22 1.26 0.2950 Microdiptera 4.30 0.80 3.20 0.57 4.25 0.84 4.50 0.99 0.50 0.6817 Total diptera 4.65 0.86 3.30 0.61 4.70 0.94 4.55 0.99 0.58 0.6274 Hemiptera Aphids 0.40 0.22 0.85 0.30 0.55 0.22 0.55 0.27 0.53 0.6655 Auchenorrhyncha 1 1.50 0.44 1.55 0.51 0.95 0.23 0.65 0.21 1.53 0.2135 Herbiv. Heteroptera 2 1.60 0.69 0.40 0.23 0.40 0.15 0.50 0.30 2.14 0.1028 Lepidoptera 0.25 0.14 0.15 0.08 0.15 0.08 0.10 0.07 0.42 0.7425 Hymenoptera Formicidae 1.10 0.56 1.60 0.72 1.50 0.84 0.85 0.50 0.27 0.8456 Macrohymenoptera 0.20 0.12 ab 0.25 0.12 a 0.00 0.00 b 0.00 0.00 b 2.33 0.0809 Microhymenoptera 0.90 0.29 0.60 0.17 0.45 0.14 0.60 0.18 0.86 0.4668 Orthoptera 1.10 0.64 0.45 0.21 0.40 0.13 0.40 0.15 0.94 0.4274 Thysanoptera 0.15 0.08 0.10 0.07 0.25 0.12 0.00 0.00 1.66 0.1839 Araneae 0.30 0.16 0.25 0.14 0.40 0.17 0.20 0.12 0.33 0.8040 Total natural enemies 3 2.50 0.84 2.70 0.88 2.35 0.97 1.65 0.56 0.30 0.8262 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ ( P 0.10) according to LSD test. No letters in row indicate no differences among means. 1 A nalysis of variance; F and P values; d. f. = 3, 73. 2 Cercopidae, Cicadellidae, Delphacidae, and Membracidae. 3 Primarily Lygaeidae and Miridae. 4 Araneae, Asilidae, Coccinellidae, Formicidae, marcohymenoptera, microhymenoptera, and Syrphidae.
43 Table 27. Effect of treatments on selected arthr opod groups (number per trap) collected in pan traps, 2008. Treatment ANOVA 1 Arthropod g roup C ontrol Bean i ntercrop 50% weeded 100% w eeded F value P Coleoptera 0.50 0.18 0.38 0.18 0.29 0.09 0.50 0.21 0.36 0.7800 Diptera Dolichopodidae 1.50 0.34 a 0.92 0.30 a 1.04 0.39 a 0.00 0.00 b 4.43 0.0060 Microdiptera 3.04 0.49 2.42 0.26 2.08 0.27 1.96 0.25 1.89 0.1362 Total diptera 4.75 0.76 a 3.33 0.41 bc 3.38 0.43 ab 1.95 0.25c 5.21 0.0023 Hemiptera Aphids 0.33 0.19 0.17 0.10 0.58 0.20 0.42 0.15 1.12 0.3456 Auchenorrhyncha 2 0.58 0.16 0.29 0.13 0.17 0.10 0.25 0.11 2.07 0.1102 Herbiv. Heteroptera 3 0.33 0.13 0.08 0.06 0.13 0.09 0.08 0.06 1.80 0.1532 Lepidoptera 0.13 0.07 ab 0.00 0.00 b 0.17 0.08 a 0.00 0.00 b 2.70 0.0506 Hymenoptera Formicidae 0.08 0.06 0.00 0.00 0.13 0.07 0.00 0.00 1.96 0.1260 Macrohymenoptera 0.21 0.12 0.21 0.10 0.21 0.08 0.25 0.11 0.04 0.9899 Microhymenoptera 1.67 0.37 a 0.67 0.19 b 1.38 0.27 ab 0.71 0.15 b 3.67 0.0151 Orthoptera 0.17 0.10 0.08 0.06 0.08 0.06 0.00 0.00 1.13 0.3413 Thysanoptera 0.46 0.12 a 0.17 0.08 b 0.17 0.08 b 0.08 0.06 b 3.59 0.0167 Araneae 0.08 0.06 0.17 0.08 0.21 0.10 0.13 0.07 0.47 0.7054 Total natural enemies 4 2.46 0.46 1.63 0.35 2.50 0.46 1.63 0.33 1.51 0.2167 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ ( P 0.10) according to LSD test. No letters in row indicate no differences among means. 1 A nalysis of variance; F and P values ; d. f. = 3,73. 2 Cercopidae, Cicadellidae, Del phacidae, and Membracidae. 3 Primarily Lygaeidae and Miridae. 4 Araneae, Asilidae, Coccinellidae, Formicidae, marcohymenoptera, microhymenoptera, and Syrphidae.
44 Table 28. Capture of selected arthropod groups (number per trap) on yellow sticky cards in weed treatments, 2007. Treatment ANOVA 1 Arthropod g roup C ontrol Bean i ntercrop 50% weeded 100% w eeded F value P Orthoptera 0.50 0.26 0.08 0.08 0.25 0.13 0.33 0.14 1.03 0.3907 Hemiptera Aleyrodidae 2.08 0.81 3.75 0.97 3.83 1.01 3.50 1.12 0.73 0.5421 Aphididae 5.75 0.84 9.92 2.05 8.83 2.58 7.67 2.25 0.73 0.5375 Auchenorrhyncha 1 6.33 1.33 7.08 1.79 7.75 1.59 4.33 0.69 1.24 0.3062 Herbiv. Heteroptera 2 0.42 0.34 0.33 0.14 0.67 0.28 0.50 0.34 0.25 0.8616 Thysanoptera 2.25 0.69 b 5.42 0.89 a 6.33 1.48 a 4.42 0.62 ab 3.07 0.0385 Coleoptera Altica spp. 0.00 0.00 0.33 0.19 0.50 0.29 1.00 0.51 1.96 0.1351 Non predatory 3 0.92 0.31 0.50 0.26 0.75 0.30 0.83 0.34 0.34 0.7970 Predatory 4 0.17 0.11 0.50 0.23 0.83 0.51 1.00 0.35 1.29 0.2906 Hymenoptera Microhymenoptera 13.17 1.86 15.33 2.09 13.67 1.53 16.58 1.51 0.76 0.5230 Macrohymenoptera 0.50 0.19 0.42 0.19 0.08 0.08 0.83 0.30 2.18 0.1049 Lepidoptera 0.58 0.19 0.42 0.15 0.50 0.19 0.58 0.26 0.15 0.9312 Diptera Dolichopodidae 4.67 0.91 6.08 0.90 4.83 0.77 5.67 1.21 0.50 0.6840 Microdiptera 27.58 4.10 23.92 4.56 25.17 2.73 26.83 3.19 0.19 0.9028 Total diptera 32.75 4.66 31.00 5.01 30.58 2.99 33.67 3.95 0.11 0.9517 Araneae 0.42 0.15 0.08 0.08 0.33 0.19 0.50 0.19 1.21 0.3178 Total predators 5 1.08 0.29 0.92 0.43 1.33 0.62 2.33 0.54 1.68 0.1853 Total natural enemies 6 14.58 1.82 16.50 2.19 15.08 1.59 19.33 1.84 1.29 0.2919 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P 0.10) according to LSD test. No letters in row indicate no differences among means. 1 Analysis of variance; F and P values ; d. f. = 3, 41. 2 Cicadellidae, Delphacidae, Membracidae, and Cercopidae. 3 Primarily Lygaeidae, and Miridae. 4 Primarily Elateridae and Scarabaeidae 5 Coccinellidae, Staphylinidae, Carabidae, Cantharidae, and Mordellidae. 6 Araneae, Predatory Coleoptera, Diptera ( Syrphidae and Asilidae), and Hemiptera (Anthrocoridae, Geocorus spp., and Reduviidae). 7 Total predators, macrohymenoptera, and microhymenoptera.
45 Table 29. Capture of selected arthropod groups (number per trap) on yellow sticky cards in weed treatments, 2008. Treatment ANOVA 1 Arthropod g roup C ontrol Bean i ntercrop 50% weeded 100% w eeded F value P Orthoptera 0.63 0.24 0.63 0.22 0.50 0.18 0.38 0.115 0.37 0.7769 Hemiptera Aleyrodidae 301.25 99.85 367.56 102.75 381.31 99.99 361.56 92.23 0.13 0.9425 Aphididae 3.88 1.03 6.81 1.90 8.13 2.27 7.75 2.14 1.03 0.3855 Auchenorrhyncha 1 7.63 1.05 a 4.69 0.43 bc 6.31 0.75 ab 3.75 0.46 c 6.30 0.0009 Herbiv. Heteroptera 2 0.31 0.18 0.63 0.22 0.75 0.27 0.38 0.15 0.96 0.4166 Thysanoptera 2.44 0.92 4.25 0.78 3.75 0.67 2.31 0.46 1.81 0.1559 Coleoptera Altica spp. 0.06 0.06 b 0.13 0.09 b 0.25 0.14 ab 0.50 0.18 a 2.29 0.0877 Non predatory 3 1.00 0.37 0.69 0.22 0.69 0.22 1.38 0.46 0.95 0.4237 Predatory 4 1.19 0.41 2.38 0.55 2.38 0.55 2.56 0.80 1.17 0.3280 Hymenoptera Microhymenoptera 35.69 4.90 35.06 2.32 38.25 2.68 32.75 2.57 0.56 0.6428 Macrohymenoptera 0.06 0.06 b 0.31 0.15 ab 0.56 0.16 a 0.31 0.12 ab 2.47 0.0709 Lepidoptera 0.13 0.09 0.00 0.00 0.19 0.14 0.31 0.12 1.70 0.1765 Diptera Dolichopodidae 8.75 1.79 10.25 1.43 9.75 1.58 6.88 1.19 0.97 0.4112 Microdiptera 14.00 2.01 15.25 2.23 15.94 1.70 17.25 1.63 0.52 0.6677 Total diptera 23.19 2.58 26.19 2.43 26.25 2.42 24.50 2.47 0.35 0.7908 Araneae 0.56 0.24 0.69 0.34 1.13 0.33 0.81 0.34 0.56 0.6407 Total predators 5 2.13 0.59 3.88 0.91 3.88 0.68 3.88 1.03 1.16 0.3344 Total natural enemies 6 37.88 4.95 39.25 2.53 42.69 2.77 36.94 3.11 0.64 0.5947 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P 0.10) according to LSD test. No letters in row indicate no differences among means 1 Analysis of variance; F and P values ; d. f. = 3, 41. 2 Cicadellidae, Delphacidae, Membracidae, and Cercopidae. 3 Primarily Lygaeidae, and Miridae. 4 Primarily Elateridae and Scarabaeidae 5 Coccinellidae, Staphylinidae, Carabidae, Cantharidae, and Mordellidae. 6 Araneae, Predatory Coleoptera, Diptera ( Syrphidae and Asilidae), and Hemiptera (Anthrocoridae, Geocorus spp., and Reduviidae). 7 Total predators, macrohymenoptera, and microhymenoptera.
46 Table 210. Effect of treatment on selected arthropod groups (number per 8 plants) observed during in situ counts, 2008. Treatment ANOVA 1 Arthropod g roup C ontrol Bean i ntercrop 50% weeded 100% w eeded F value P Orthoptera 0.79 0.23 0.63 0.13 0.54 0.13 0.33 0.16 1.50 0.2198 Hemiptera Aphididae 0.00 0.00 0.38 0.26 0.13 0.09 0.83 0.83 0.71 0.5504 Auchenorrhyncha 2 0.54 0.18 0.25 0.12 0.21 0.10 0.17 0.08 1.86 0.1415 Herbiv. Heteroptera 3 0.25 0.15 0.13 0.09 0.38 0.17 0.63 0.46 0.66 0.5789 Coleoptera Altica spp. 0.00 0.00 b 0.04 0.04 b 0.04 0.04 b 0.29 0.14 a 3.1 0.0307 Non predatory 0.13 0.07 0.04 0.04 0.13 0.07 0.13 0.07 0.43 0.732 Predatory 0.04 0.04 0.00 0.00 0.00 0.00 0.04 0.04 0.66 0.5793 Hymenoptera Formicidae 0.25 0.17 0.75 0.20 0.88 0.35 1.13 0.40 1.51 0.2168 Microhymenoptera 0.21 0.10 0.33 0.16 0.33 0.13 0.13 0.17 0.89 0.4496 Lepidoptera Larvae 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.00 1.0000 Larvae + adults 0.08 0.06 0.08 0.06 0.04 0.04 0.17 0.10 0.62 0.6021 Diptera Dolichopodidae 0.88 0.35 1.38 0.39 0.96 0.24 0.75 0.26 0.72 0.5405 Microdiptera 1.33 0.39 b 3.42 0.82 a 3.96 0.90 a 1.54 0.28 b 4.04 0.0097 Total diptera 2.29 0.53 b 4.88 0.97 a 5.00 1.02 a 2.29 0.44 b 3.77 0.0134 Aranae 1.67 0.39 2.38 0.41 2.29 0.40 2.21 0.37 0.64 0.5891 Total predators 4 2.00 0.42 3.17 0.40 3.17 0.42 3.42 0.53 0.66 0.1124 Total natural enemies 5 2.25 0.44 3.50 0.36 3.50 0.41 3.54 0.54 2.07 0.1104 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P 0.10) according to LSD test. No letters in row indicate no differences among means. 1 Analysis of variance; F and P values; d. f. = 3, 89. 2 Primarily Cicadellidae. 3 Primarily Arvelius albopunctatus (Pentatomidae) 4 Araneae, Asilidae, Coccinellida, and Formicidae. 5 Total predators macrohymenoptera, and microhymenoptera.
47 Table 211. Effect of weed treatments on pepper yields (number fruit per plot ), 2007. Treatment Fancy 1 U.S. No. 1 U.S. No. 2 Unweeded Control 0 0.00 0.00 2.75 2.43 c Bean Intercrop 0 0.00 0.00 15.25 4.96 ab 50% Weeded 0 1.50 1.50 12.50 2.66 bc 100% Weeded 0 0.75 0.36 23.25 3.68 a ANOVA 2 F value 0 1.00 7.44 P 0 0.4363 0.0083 Data are means standard errors of 4 replications. Means in columns followed by the same letter do not differ (P 0.10) according to LSD test. No letters in column indicate no differences among means. 1 Pepper fruit grades defined in text. 2 Analysis of variance; F and P values; d. f. = 3, 9. Table 212. Effect of weed treatments on pepper yields (number fruit per plot ), 2008. Treatment Fancy 1 U.S. No. 1 U.S. No. 2 Unweeded Control 0.78 4.34 b 5.97 2.85 c 14.62 2.42 Bean Intercrop 8.40 2.87 a 23.41 4.12 ab 27.80 7.58 50% Weeded 3.75 0.63 ab 16.75 3.73 b 21.75 3.97 100% Weeded 9.04 0.50 a 29.06 3.22 a 17.79 3.25 ANOVA 2 F value 3.49 12.67 1.39 P 0.0633 0.0014 0.3075 Data are means standard errors of 4 replications. Means in columns followed by the same letter do not differ (P .10) according to LSD test. No letters in column indicate no differences among means. 1 Pepper fruit grades defined in text. 2 Analysis of variance; F and P values; d. f. = 3, 9.
48 CHAPTER 3 THE IMPACT O F INTERCROPPING SQUASH WITH NONC ROP VEGETATION BORDERS ON THE ABOVEGROUND ARTHROPOD COM MUNITY Introduction Non crop vegetation within the landscape has enhanced populations of predators and parasitoids in several agricultural ecosystems (Landis et al. 2005, Denys and Tscharntke 2002, Andow 1991) by providing resources including pollen, nectar, alternative prey, shelter, and overwintering habitat ( Altieri and Nichols 2004 a Denys and Tscharntke 2002, Landis et al 2005, van Emden 1965). However, mechanized agriculture requires planting in rows, so that integration of noncrop vegetation into the agricultural landscape could be accomplished most easily as long strips or rows. Rows of noncrop vegetation can provide several ecosystem services, impact arthropod population dynamics (Nentwig 1988) and may also be responsible for a reduction in pest outbreaks ( Altieri and Nichols 2004, Pimental 1961) Integrating rows of noncrop vegetation into the agricultural landscape may influence arthropod populations in adjacent crops as well (Landis et al. 2000, Showler and Greenburg 2002) When grassy corridors were interplanted within soybean fields, the distribution of colonizing potato leaf hoppers, Empoasca fabae (Harris), was disrupted such that yields in intercropped plots were within 2% of thos e in insecticide treated plots lacking corridors (Kemp et al. 1989). Weedy strips incorporated into cotton reduced pest pressure and increased natural enemy populations (Showler and Greenburg 2002). Similarly, augmenting vegetable production systems with i nsectary hedgerows consisting of a strip of flowering plants (Pascual Villalobos et al. 2006) or weeds (Bugg 1987) has increased natur al enemy activity in neighboring vegetable crops.
49 Intercropping provides vegetational camouflage that confounds insect visual orientation to a host plant and provides a bouquet of chemical volatiles that interfere with olfactory driven host finding mechanisms (Smith and McSorley 2001). The a dditional vegetational architecture may also serve as a physical barrier t hat interferes with the dispersal patterns of weak flying arthropods (Compton 2002, Lewis 1969) or those that employ aerial drift for dispersal (Bonte 2008). Additionally, the establishment of a border crop as a protective barrier around a main crop has r educed insect vectored plant viruses (Murphy 2008, Damicone et al. 2006, Fereres 2000). In the current study, we examine the influence of four different border treatments on the aboveground arthropod community in a diverse agricultural system. The borders consisted of plants or fallow treatments arranged in rows on either side of a vegetable crop. By using various sampling techniques, the objectives were to determine whether border crops function as barriers to or colonization sites for pests and natur al enemies, to assess the arthropod groups associated with each border treatment, and to evaluate the effect of border crop composition on the arthropod populations in an adjacent squash ( Cucurbita pepo L.) crop. Materials and Methods Field Management and Experimental Design Field experiments were conducted during the fall growing seasons in 2007 and 2008 at the University of Florida Plant Science Research and Education Unit (P S R E U) (29o24N, 82o9W) located in Citra FL. The soil in the site is classified as Arredondo sand (95% sand, 2% silt, 3% clay) (Thomas et al. 1979) with 1.5% organic matter. Prior to establishing the experiment and between growing seasons, fields were maintained in weedy fallow. In May of 2007, Fusilade (Syngenta Crop Protection, Inc., Greensboro,
50 NC) was applied over the experimental area to control annual and perennial grasses and encourage the establishment of broadleaf weeds that may provide resources for insect her bivores and natural enemies. Four treatments bordering a cultivated squash crop were assessed: pigeon pea ( Cajanus cajan [L.] Millsp.), Vegetable; Sorghum sudangrass x sudangrass hybrid (Sorghum bicolor Moench S sudanense [Piper] Stapf), g rowers ch oice, the native weed complex, and a bare ground control. B order treatments were established during the summer (5 July 2007 and 9 June 2008) prior to planting squash to ensure that border plants would be taller than the crop at the time of transplanting. Pigeon pea was handseeded 0.3 m apart in four rows with 0.6 m spacing. The sorghum sudangrass hybrid was drilled mechanically ( Sukup Manufacturing Company, Sheffield, I A ) at 27 kg/ha. The native weed complex was permitted to grow undisturbed during the growing season. The composition of weed treatment plot was evaluated by sampling weeds from two 0.5 m2 q uadrats one week prior to planting squash. Weeds were categorized as broadleaf weeds, grasses or sedges, and identified at the genus and species level when possible. The aboveground biomass was calculated by cutting weeds at the ground level, and drying it at 70 S ome weeds grew among the pigeon pea plants ; therefore, weed biomass was calculated using the same method for pigeon pea treatment at the end of the season. The bare fallow treatment was maintained manually using a push rototiller ( Pro line FRT Garden Tiller, Troy Bilt Products, Cleveland, OH) and hand weeding. Plots were 8 m x 7.2 m wit h 3 m spacing between plots and arranged in a randomized complete block design with four replications. Experimental blocks were separated by a 3 m wide strip of undisturbed perennial weed reservoir running the
51 entire length of the block. These uncultivat ed strips w ere comprised primarily (<80% by biomass) the following broadleaf weeds: Florida pusley ( Richardia scabra L.), h orseweed ( Conyza canadensis [ L .] Cronquist ), cutleaf eveningprimrose ( Oenothera laciniata Hill), Mexican tea (Chenopodium ambrosioid es L. ), and Sida sp All rows were orientated east to west to reduce the effects of shading of squash by border plants. Border treatments measuring 8 m x 1.8 m bordered both sides of the plot with a 4.3 m x 3.6 m squash planting centered in between. Thr ee rows of the yellow squash cultivar Early Crookneck with 0.9 m spacing were handtransplanted (18 September 2007 and 14 October 2008) into the various plots The threeweek old transplants were placed 0.4 m apart, resulting in a density of 36 plants per plot. All weeding was done manually and squash plants were fertigated (16.8 k g N/ha) using driptape. Space between plots was cultivated manually and using a pushrototiller. Height and plant stand (plants/m) of border crops were measured at the time of squash transplanting and approximately one month later. A section of the border (1 m) was selected on each side of the plot, the number of plants per meter were counted and averaged together as plant stand for pigeon pea and sorghum sudangrass. The number of squash plants per plot was assessed and percent defoliation due to insect injury was calculated weekly based on the following scale, <25%, 2550%, 5199% defoliated, or missing. Arthropod Sampl ing Arthropod communities were sampled within both the border treatments and squash planting using three sampling techniques: sticky card traps, pitfall traps, and pan traps. Traps were set in the morning before noon and left in the field for 24 h; the d ay of collection was recorded as the sampling date. T axa collected in traps were identified to family level or guild and counted.
52 Sticky Cards. A yellow, unbaited, Pherocon AM trap (Great Lakes IPM Vestaburg, MI) was placed 5 cm above the ground between squash rows and within one border in each plot. Traps were set biweekly during both the 2007 and 2008 growing seasons beginning the week of transplanting. Collected sticky cards were wrapped in plastic food wrap (Stretchtite, Polyvinyl Films Inc., Su tton, MA) and stored at 4 Insects trapped within the grid (23 cm x 18 cm) on the sticky card were counted and recorded. A representative sample of the whiteflies and thrips on each card were counted using an exclusionary grid that allows one to count 15 of the 63 grid squares (2.5 cm x 2.5 cm) on each card (Finn 2002). Pitfall Traps. Plastic sandwich containers (14 cm x 14 cm x 4 cm; 532 mL) were used as pitfall traps (Triplehorn and Johnson 2005) and buried so that the upper edge was flush with soil surface. The traps were filled three quarters (ca. 300 mL) with water, along with 3 to 4 drops of dish detergent (Ultra Joy, Procter & Gamble, Cincinnati, OH) to the break surface tension and prevent escape. Two pitfall traps one in the border and one between rows of squash, were set in each plot every three weeks beginning the week of transplanting Pan Traps. Clear polyethylene deli containers (11 cm in diameter x 4.5 cm deep; 236 mL) (Gainesville Paper Company, Gainesville, FL) were used as pan traps. Traps were placed at midplant height between squash rows and at the same height in one of the borders, and filled half way with water (ca. 175mL) along with 2 to 3 drops of dish detergent (Southwood and Henderson 2000). Pan traps were set biweekly thro ughout the growing season beginning the first week after transplanting. In situ counts. Whole plant visual counts were performed weekly throughout both growing seasons by systematically selecting four plants from each plot. All of the
53 leaves were examined and arthropods were identified to order or family in the field, and counted. When present, key pests were identified to genus and species. Statistical Analysis Arthropods collected in traps were identified to order and at the family level. In many cas es, orders represented by only a few individuals in several families were grouped together by order or feeding guild for statistical analysis. Commonly occurring arthropod groups and one particularly frequent genus were analyzed. Data for all members of the entire arthropod community were reported for each of the sampling methods. However, pitfall trapping targets organisms found at the soil surface, while pan trap and stickycards are typically used to sample flying insects (Southwood and Henderson 2000) Data from arthropod counts were analyzed by repeated measures analysis of variance (ANOVA) using the GLM procedure (SAS Institute 2008). Means were separated with the least significant difference (LSD) test at P 2008). Pitfall trap data for Formicidae were log transformed by log10 (x+1) prior to analysis to stabilize variance, but untransformed means are reported. Results Border Composition At the time of transplanting, heights of plants in borders were measured and plant density was established. Height and plant stand of border treatment s are show n in ( Table 3 1). The total rainfall at the Citra research station during the first 3 weeks of border treatment establishment was 3.3 cm ( FAW N 2009). Some of the sorghum sudangrass and pigeon peas succumbed to drought conditions resulting in a somewhat lower plant density than expected. During the 2008 growing season, the total rainfall
54 during the 3week period after planting was 17.9 cm ( FAW N 2009 ) which led to more densely established pigeon pea and sorghum sudangrass. An understory of lowgrowing weeds dominated by Florida pusley ( R. scabra) crowfoot grass ( Dactyloctenium aegyptium [L.] Willd ), Bermuda grass ( Cyndon dactylon L.), crabgrass ( Digitaria spp.), and sedges grew in the pigeon pea treatment. The native weed complex was composed of a mixture of broadleaf weeds, grasses, and sedges. During 2007, the native weed complex was nearly 60% (biomass) grasses, namely barnyardgrass ( Echin ochloa crus galli (L.) Beauv ) and crabgrass ( Digitaria spp.). Broadleaf weeds, primarily Florida pusley ( R. scabra) and hairy indigo ( Indigofera hirsuta L ) comprised about 30% of the native weed complex, while the remainder were sedges ( Cyperus spp.). Broadleaf weeds dominated more than 50% of the native weed complex during the 2008 growing season, while grasses comprised 33%, and sedges the remai ning 16%. Arthropod Sampling Border 2007. Most differences within the border treatments were detected using sticky cards (Tables 2 2; 23). Total natural enemies including, predatory Diptera, Coleoptera, Hemiptera, Hymenoptera, parasitoids, and Aranae (spiders), were higher on sticky cards and in pitfall traps in the native weed complex and pigeon pea treatment s when compared to sorghum sudangrass ( Table 32). Although the total natural enemies were lower ( P = 0.0351) on sticky cards placed in sorghum sudangrass, predat ory Coleoptera, which included primarily the famil y Coccinellidae followed by Staphylinidae, Carabidae, Mordellidae, and Cantharidae, wer e more abundant ( P = 0.0166) in sorghum sudangrass than in any other treatment. Dolichopodidae (longlegged flies) were found in lowest numbers in the sorghum -
55 sudangrass treatment on sticky cards as well. Microhymenoptera were most abundant ( P = 0.0119) in the pigeon pea and weeds on sticky cards and in weeds in pitfall traps. The most ( P = 0.0001) spiders were trapped on sticky cards in the bare ground control. Among the pest insects captured on sticky cards, Orthoptera ( grasshoppers and crickets) were most abundant ( P = 0.0048) in the native weed complex and pigeon pea ( Table 33). Both Thysanoptera ( thrips ) mainly Frankliniella spp. ( P = 0.0006) and fleabeetl e s, Altica spp., ( P = 0.0022) were detected in highest numbers in the bare ground control. Significant differences were not observed for any Hemiptera groups between the native weed complex and the control on sticky cards. However, when those treatments were compared to sorghum sudangrass, fewer Aphididae ( aphids ) ( P = 0.0049), C icadellid ae ( leaf hoppers ) ( P = 0.0425), and Aleyrodidae ( whiteflies ) were found than in bare ground. Membracidae (treehoppers) were more abundant ( P = 0.0239) in pigeon pea than sorghum sudangrass as well. When pan traps were used for sampling, auchenorrhynchan activity was higher in the native weed complex than in sorghum sudangrass and the control ( Table 33). Microdiptera and total D iptera were more numerous in the weeds than in the sorghum sudangrass as well. Aphididae were more abundant in pan traps placed in Sorghum sudangrass than in the control, opposite from the results observed with sticky cards. In pitfall trap captures, more nonpredatory Coleoptera (primaril y Altica spp. and Elateridae) were trapped in the bare ground control than in pigeon pea or sorghum sudangrass ( Table 33). Although several arthropods groups were impacted by field border treatments, no differences were observed for Lepidoptera or herbi vorous Heteroptera using any of the sampling methods ( Table 33).
56 Border 2008. Several of the trends observed in 2007 were repeated during 2008. The groups microhymenoptera ( P = 0.0152), D olichopodidae ( P = 0.0011), and total natural enemies ( P = 0.0 157) were most common on sticky cards placed in pigeon pea; however most (P = 0. 0119) of the spiders were trapped in the control on sticky cards or pan traps ( Table 34). In pan traps, microhymenoptera numbers were higher ( P = 0.0275) in the native weed com plex and sorghum sudangrass than in the control. More total natural enemies were collected in pitfall traps in the native weed complex than the bare ground control but no differences were observed with the other sampling methods. During 2008, significa ntly more ( P = 0.0055) auchenorrhynchans were captured on sticky cards and pan traps in the native weed complex and pigeon pea than the control ( Table 35). The native weed complex had the most ( P = 0.0211) cicadellid activity, while membracids were most abundant ( P = 0.0 004) in pigeon pea. Crop 2007 Fewer differences existed among treatments when sampling occurred within the crop, and no differences were measured in pitfall trap captures during 2007 ( Table 36, 37). The most spiders were trapped in s quash bordered by sorghum sudangrass using sticky cards or pan traps ( Table 36), however no other beneficial arthropod collected within the crop was impacted by the border treatment. Auchenorrhynchans were more abundant on sticky cards in squash border ed by the native weed complex than by sorghum sudangrass ( Table 37). Additionally, Altica spp. were least abundant when sorghum sudangrass and weeds bordered the squash. When in situ counts were performed, ants were more numerous ( F = 3.09; d. f. = 3, 41; P = 0.0376) on squash bordered by bare ground (10.08 2.67) treatments than on squash bordered by sorghum sudangrass (4.08 1.85) and weeds (3.33 1.19). In situ counts also revealed melonworm ( D iaphania hyalinata L.) and saltmarsh ( Estigmene
57 ac rea [Drury]) caterpillars present on many plants. Their numbers were not affected by the border treatments, but melonworms, which were more common than saltmarsh caterpillars, averaged 5.89 1.37 caterpillars per plant over the season, reaching a high of 15.15 3.87 caterpillars per plant on 30 Oct. 2007. These numbers resulted in heavy defoliation and plant mortality with 97.06 4.33% of the plants in each plot more than 50% defoliated. Therefore, yield data could not be collected. Regardless of sampling method, several other groups were unaffected by the presence of border treatments including Lepidoptera, herbivorous Heteroptera Coleoptera, and beneficial insects from families within Coleoptera, Diptera, Hymenoptera, and Hemiptera. Crop 2008 In 2008, more predatory Coleoptera, mainly coccinellids were trapped on yellow sticky cards in squash bordered by sorghum sudangrass, than by p igeon pea or the bare ground control ( Table 38). Cicadellids were more abundant in squash bordered by weeds than by pigeon pea or the control (Table 39) Pigeon pea bordered squash had more membracids than squash bordered by sorghum sudangrass or the native weed complex (Table 39) In pitfall traps, total auchenorrhynchans were more numerous in squash bordered by w eeds than by sorghum sudangrass or bare ground. During in situ counts, auchenorrhynchans were only ( F = 4.53; d. f. = 3, 73; P = 0.0057) observed on squash bordered by the native weed complex (0.45 0.21). During in situ counts, the most (F = 4.97; d. f. = 3, 73; P = 0.0034) aphids were observed on squash plants bordered by the bare ground control (112.65 22.49 aphids per plant ) compared to sorghum sudangrass (60.35 12.91), pigeon pea (44.95 11.45), and the native weed complex (40.45 8.28). Howev er, aphid captures using
58 sticky cards and pan traps did not differ. No differences were observed for any groups in pan trap captures during 2008. As in the previous season, melonworms and saltmarsh caterpillars were present, but did not differ among treatments. Melonworms were the dominant lepidopteran species averaging 2.56 0.61 caterpillars per plant over the season. By 18 November 45.13 12.29 % of the plants in each plot were more than 50% defoliated. The crop was further damaged by a freeze on 20 November whe n temperatures dropped to 1.81 C (FAWN 2009). This m arked the end of the experiment and prevented yield data from being collected. Discussion The border treatments in this study varied greatly in terms of the arthropods they hosted, due in part to the resources provided by the borders. The sorghum sudangrass treatment was both the tallest and most densely planted border treatment. Based on hei ght and planting density, sorghum sudangrass had more green surface area to serve as habitat for arthropods. However, sorghum sudangrass often had the lowest populations of pests and natural enemy species. Few differences were observed between the pigeon pea treatment and the native weed complex which may have been due to the weedy understory that developed over the course of the season beneath the pigeon pea. This research provided data for several objectives. The first question we hoped to address w as the impact of border treatments on the migration of ambient flying pests into the squash crop. The bare ground control provided an unobstructed highway between the perennial weed refugia and squash crops. Notably, more weak flying insects, including w hiteflies (Basu 1995) and thrips (Lewis 1997) as well as spiders that rely on aerial drift for dispersal (B onte et al. 2008) were trapped on sticky cards in the
59 bare ground control than in the sorghum sudangrass treatment. However, the border treatments ultimately had no impact on the movement of whiteflies, a key pest of squash in north Florida, or thrips into the squash crop. I n experiments, conducted in Oklahoma, squash plants intercropped with sorghum ( Sorghum bicolor ( L. ) Moench) had reduced inciden ce of aphidborne viruses (Damicone et al 2006), even though alate captures in modified pan traps were not different. No treatment differences in aphids captures by sticky cards and pan traps placed with in the squash crop wer e observed in the current study. Fewer aphids were recorded on sticky cards in sorghum sudangrass compared to the control and weedy fallow in 2007. In addition, fewer aphids were found on squash plants bordered by sorghum sudangrass during in situ counts in 2008. This suggests that during the 2008 season, sorghum sudangrass may have functioned as a barrier to aphid dispersal and colonization. Two possibl e explanations may account for the reduced colonization or dispersal of aphids in squash bordered by a sorghum sudangrass treatment. First, sorghumsudangrass may act as a direct physical barrier preventing aphids from entering the field, or it may alter the background of host crop/vegetation and subsequently reduce the potential for aphid alightment. Thus, aphids maybe investing m ore time and energy probing a diversionary intercrop, rather than attacking the main crop (Smith and McSorley 2000). Trenbath (1977) describes this phenomenon as the fly paper effect because herbivores are "lost to plant surfaces other than the main cr op, which may delay colonization and increase their exposure to mortality factors (Smith and McSorley 2000, Trenbath 1993, 1977). Secondly, the high population of coccinellids may have increased the potential for enhanced natural control, which will ultimately reduce aphid numbers. It should be noted that yellow sticky cards serve as an attractant for aphids (Southwood and
60 Henderson 2000) ; therefore, further testing using clear plexiglass sticky boards, or other passive sampling methods may offer a better understanding of natural patterns of dispersal and movement of aphids in this system (Powell et al.1998). Our second objective was to observe and document the components of the aboveground arthropod community within each of the border crop treatments. Each treatment provided a unique habitat that was attractive to particular arthropods while most other groups remained unaffected by the border composition. Natural enemies were generally highest in the pigeon pea and native weed complex borders which were t he most ve getationally diverse treatments. However, sorghum sudangrass had the highest number of predatory Coleoptera (sticky cards) during both seasons Spiders were most abundant in the bare ground fallow on sticky cards and sometimes pan traps rather than in pitfall traps which measure activity at the level of the soil surface. Since uncultivated strips in the agricultural landscape have been demonstrated as an important source of spiders (Netwig 1998), it is possible that aerial dispersal of spiders from the adjacent weed refug ia blew through plots unobstructed by border treatments more readily than when any of the other treatments were there. Non predatory Coleoptera were unaffected by the composition of the border treatment with the exception of t he herbivorous flea beetle, Altica spp. (Chrysomelidae). Fleabeetles were most often captured on sticky cards in the bare ground control, while they were rarely recovered in pan traps and pitfall traps. Tahvanianen and Root (1972) demonstrated that inter cropping interfered with host finding and feeding behavior of the flea beetle, Phyllotreta cruciferae (Goeze). The habitat simplicity that results from cultivation and the absence of a diverse background may signal favorable conditions to
61 colonizing insec ts like flea beetles (Root 1973, Cromartie 1975, Dosdall et al. 1999), whose host finding behavior appears to be adapted to bare soil (Cromartie 1975). The third objective of this study was to evaluate whether planting borders impacted arthropod populati ons in the adjacent crop, providing a spill over effect. Although total natural enemies were most abundant in the native weed complex no corresponding increase in total natural enemy captures was detected in the adjacent squash. This may be the result of lag time or attributed to the type of border crops employed. However, in 2008 greater numbers of predatory Coleo ptera trapped on sticky cards in sorghum sudangrass corresponded to the higher populations trapped within the squash crop. In 2007, severe squash defoliation by melonworm may have prevented significant numbers of predatory Coleoptera from moving into the squash crop. The spill over effect was also observed to a limited extent for auc henorrhyncans, although the trends were not consistent between seasons. Since leafhoppers are not a major pest of squash, the impact of high leafhopper populations in squash was negligible. Because this was a community study, multiple methods were employed to assess arthropod populations. Not every method was best suited for every group, thus discrepancies between results for the same arthropod group using different sampling tools are to be expected. The use of yellow sticky cards is a standard sampling method to determine relative abundance (Southwood and Henderson 2000). However, colored traps do not provide information regarding the natural movement patterns of insects (Powell et al. 1996). Passive traps including clear pan traps and pitfall traps likely provide more information regarding natural patterns of locomotion (Powell et al. 1996; Southwood and Henderson 2000). However in these experiments pan traps captured
62 fewer arthropods per trap than pitfall and sticky card traps left in the field for the same period of time. Although some differences were det ected using pan traps, increasing the time that they are placed in the field m ay provide more data. Crop diversity is thought to have a stabilizing effect on arthropod populations and reduce pest outbreaks (Pimental 1961). However, the intercropped refuge strips in this study w ere not sufficient to prevent pest outbreak This was most evident from in situ counts Melonworm, a specialist herbivore considered to be a secondary pest of squash in Florida, was not hindered by border crops and was responsible for nearly 100% crop loss by the end of the 2007 and nearly 50% crop loss 2008. Generalist defoliators like the saltmarsh caterpillar appeared unencumbered by the border treatments as well. Although crop diversity impacts a wide range of pest and beneficial species (Landis et al. 2000, Andow 1991), many of the studies examined in these reviews included only a few key pests or natural enemies rather than the full arthropod community. In the current study, the impact of veget ational diversity on the arthropod community was evaluated, and while a number of members responded, most groups were unaffected by treatments. These results occurred even with a fairly liberal criterion ( P 0.10) selected to detect differences among treatments These results are interesting in that they are consistent with the possibility that a large number of arthropod groups may be unaffected by vegetational diversity at least under the current conditions Uncultivated weed refugia were established between each of the experimental blocks to provide a source of arthropods that could migrate into the border treatment areas and the squash crop (van Emden 1965) The presence of weed refugia mimicks
63 the reality of low resource and organic farmers who are typically forced to tolerate some level of weeds because of limited weed control options or labor constraints (Brberi 2002). Since plant diversity in the habitat surrounding a crop may have stronger affects on the abundance of certain speci es than the host plant patch size (Bach 1984), plot dimensions and the proximity of weed refugia may be responsible for attenuated treatment affects, especially in the bare fallow treatment. The results of the current experiment suggest s that incorporating strips of pigeon pea, sorghum, or weeds may not make a significant contribution to the activity of natural enemies in neighboring crops in a site that already has a considerable amount of crop or noncrop diversity in the adjacent landscape. Results cannot be generalized to conventional vegetable production systems that operate against a less diverse background because these systems may lack reservoirs from which resident arthropods could move to colonize intercropping strips or crops.
64 Table 31. Border treatment mean height and plant stand one week prior to trans planting and one month after transplanting. Treatment Pre -transplant Post-transplant Height (cm) Stand (plants/m) Height (cm) Stand (plants/m) 2007 Bare ground 0 0 0 0 Pigeonpea 80.94 4.81 98.81 4.88 Sorghum 114.19 48.25 115.19 50.75 Weeds 44.81 0 41.88 0 2008 Bare ground 0 0 0 0 Pigeonpea 87.38 5.88 79.75 4.50 Sorghum 116.06 62.88 91.31 61.63 Weeds 42 0 30.63 0
65 Table 32 S elected beneficial arthropod groups (number per trap) recovered on sticky card s, pan traps and pitfall traps within border treatments, 2007. Treatments ANOVA 1 Arthropod group Trap Bare ground Pigeon pea Sorghum Weeds F P Coleoptera Predatory 2 Sticky 1.00 0.33 b 0.67 0.26 b 3.00 1.06 a 0.50 0.29 b 3.83 0.0166 Pitfall 1.00 0.30 1.33 0.47 0.42 0.19 1.83 0.75 1.50 0.2290 Diptera Dolichopodidae Sticky 5.58 1.21 a 5.67 0.97 a 0.83 0.27 b 3.58 0.91 a 6.21 0.0014 Pan 0.33 0.19 0.83 0.42 0.00 0.00 0.58 0.26 1.71 0.1805 Pitfall 2.42 0.69 2.83 1.02 0.58 0.29 2.50 0.92 1.60 0.2051 Hymenoptera Formicidae Sticky 0.25 0.13 0.50 0.26 0.92 0.43 0.42 0.19 1.02 0.3943 Pan 0.17 0.11 b 0.75 0.31 b 2.50 0.93 a 0.25 0.18 b 2.62 0.0303 Pitfall 27.18 1.23 25.92 1.38 25.92 1.29 20.88 1.38 0.15 0.9273 Microhymenoptera Sticky 16.92 2.38 b 29.25 2.38 a 16.33 3.46 b 30.92 6.35 a 4.13 0.0119 Pan 0.42 0.15 0.33 0.14 0.33 0.14 0.50 0.20 0.27 0.8481 Pitfall 1.75 0.72 b 2.17 0.68 b 1.25 0.43 b 4.42 0.97 a 3.74 0.0183 Araneae Sticky 1.58 0.37 a 0.58 0.26 b 0.00 0.00 b 0.17 0.11 b 9.93 0.0001 Pan 0.08 0.08 0.08 0.08 0.25 0.13 0.08 0.08 0.75 0.5312 Pitfall 1.08 0.31 1.08 0.31 1.17 0.34 2.08 0.75 1.01 0.3981 Total natural enemies 3 Sticky 25.75 2.82 ab 37.42 4.03 a 22.00 3.70 b 36.08 7.27 a 3.15 0.0351 Pan 1.00 0.28 2.25 0.51 3.08 1.04 1.58 0.53 1.80 0.1615 Pitfall 6.75 1.04 bc 8.25 1.36 ab 3.92 0.58 c 11.67 1.95 a 6.00 0.0017 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P ding to LSD test. No letters in rows indicate no differences among means. 1 Analysis of variance; F and P values; Sticky cards (d. f. = 3, 41); Pan traps (d. f. = 3, 41); Pitfall traps (d. f. = 3, 41). 2 Coccinellidae, Staphylinidae, Carabidae, Cantharidae, and Mordellidae. Predatory Coleoptera were rarely observed in pan trap s, and were included in Total natural enemies. 3 Aranae, Formicidae, Mutillidae, microhymenoptera, macrohymenoptera, predatory Coleoptera, Diptera (Asilidae, Bombyliidae, Dol ichopodidae and Syrphidae), and Hemiptera ( Geocoris spp., Anthrocoridae, and Reduviidae).
66 Table 3 3 Selected arthropod groups (number per trap) recovered on sticky card s, pan traps and pitfall traps within border treatments, 2007. Treatments ANOVA 1 Arthropod group Trap Bare ground Pigeon pea Sorghum Weeds F P Orthoptera Sticky 0.08 0.08 b 0.83 0.27 a 0.00 0.00 b 1.00 0.35 a 5.00 0.0048 Pan 0.17 0.11 0.00 0.00 0.25 0.13 0.08 0.08 1.24 0.3067 Pitfall 1.33 0.50 1.75 0.37 3.00 2.12 2.17 0.60 0.39 0.7633 Hemiptera Aleyrodidae Sticky 8.33 2.73 a 5.67 1.97 ab 1.58 1.23 b 3.33 0.96 ab 2.51 0.0720 Aphididae Sticky 15.33 3.82 ab 12.17 2.74 bc 1.33 0.41 c 25.75 7.62 a 4.98 0.0049 Pan 0.17 0.11 b 0.42 0.15 ab 0.92 0.38 a 0.17 0.11 b 2.49 0.0734 Pitfall 0.17 0.11 1.08 0.43 0.83 0.32 0.75 0.35 1.32 0.2809 Auchenorrhyncha 2 Sticky 9.03 1.98 a 6.67 1.39 ab 3.67 0.50 b 8.25 1.47 ab 3.22 0.0333 Pan 0.25 0.25 b 1.17 0.49 ab 0.83 0.24 b 4.08 1.94 a 2.75 0.0551 Pitfall 2.75 0.63 3.75 0.64 2.67 0.91 4.58 1.31 0.95 0.4270 Cicadellidae Sticky 7.33 1.48 a 3.83 0.92 bc 3.58 0.62 c 7.08 0.14 ab 2.98 0.0425 Membracidae Sticky 1.00 3.25 ab 1.67 0.55 a 0.00 0.00 b 0.75 0.37 ab 3.50 0.0239 Heteroptera 3 Sticky 0.75 0.43 0.33 0.19 0.08 0.08 0.25 0.13 1.37 0.2649 Pan 0.00 0.00 0.08 0.08 0.25 0.18 0.25 0.13 1.07 0.3725 Pitfall 0.50 0.23 0.50 0.20 0.25 0.18 0.92 0.38 1.73 0.1763 Coleoptera Altica spp. Sticky 5.08 2.06 a 0.08 0.08 b 0.00 0.00 b 0.00 0.00 b 5.74 0.0022 Pan 0.08 0.08 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.4025 Non predatory 4 Sticky 1.08 0.25 1.08 0.29 0.75 0.31 0.25 0.13 1.96 0.1350 Pan 0.17 0.11 0.00 0.00 0.17 0.11 0.25 0.13 1.08 0.3693 Pitfall 083 0.27 0.33 0.19 0.17 0.11 0.50 36 1.35 0.2723 Lepidoptera Sticky 0.33 0.26 0.17 0.11 0.50 0.26 0.42 0.26 0.41 0.7452 Pan 0.25 0.13 0.25 0.18 0.17 0.17 0.17 0.11 0.10 0.9580 Pitfall 0.25 0.13 0.75 0.28 0.75 0.22 0.25 0.13 2.12 0.1124
67 Table 3 3. Diptera Microdiptera Sticky 14.58 4.90 9.08 1.27 11.33 3.26 7.83 1.62 0.89 0.4543 Pan 1.33 0.36 b 1.67 0.53 ab 0.83 0.51 b 3.08 0.97 a 2.79 0.0526 Pitfall 0.67 0.36 2.00 0.60 1.42 0.57 2.17 0.74 1.41 0.2536 Thysanoptera 5 Sticky 9.83 2.23 a 3.67 0.86 b 0.33 0.19 b 4.25 1.66 b 7.07 0.0006 Pan 0.00 0.00 0.17 0.17 0.00 0.00 0.00 0.00 1.00 0.4025 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P ding to LSD test. No letters in rows indicate no differences among means. 1 Analysis of variance; F and P values; Sticky cards (d. f. = 3, 41); Pan traps (d. f. = 3, 41); Pitfall traps (d. f. = 3, 41). 2 Cercopidae, Cicadellidae, Delphacidae, and Membracidae. 3 Herbivorous Aleyidae, Lygaeidae, Miridae, Pentatomidae, Pyrrochoridae, and Rhopalidae. 4 Chrysomelidae, Elateridae, Tenebrionidae 5 Primarily Thripidae that were rarely observed in pitfall traps.
68 Table 34 Selected beneficial arthropod groups (number per trap) recovered on sticky card s, pan traps and pitfall traps within bor der treatments 2008. Treatments ANOVA 1 Arthropod group Trap Bare ground Pigeon pea Sorghum Weeds F P Coleoptera Predatory 2 Sticky 0.42 0.19 b 0.25 0.18 b 1.42 0.48 a 0.42 0.19 b 3.45 0.0251 Pitfall 0.08 0.08 0.42 0.15 0.08 0.08 0.17 0.11 1.50 0.2290 Diptera Dolichopodidae Sticky 1.25 0.31 b 3.92 0.50 a 0.92 0.29 b 1.67 0.80 b 6.49 0.0011 Pan 0.25 0.11 0.75 0.31 0.38 0.16 0.21 0.10 2.04 0.1189 Pitfall 0.33 0.23 ab 0.17 0.11 b 0.00 0.00 b 0.83 0.32 a 2.92 0.0453 Hymenoptera Formicidae Sticky 0.25 0.13 a 0.00 0.00 b 0.00 0.00 b 0.00 0.00 b 3.80 0.0170 Pan 0.19 0.14 2.00 0.70 0.25 0.11 2.00 1.74 1.16 0.3347 Pitfall 11.91 7.91 b 25.58 5.30 ab 24.67 8.07 ab 36.92 9.64 a 2.25 0.0967 Microhymenoptera Sticky 7.75 1.01 b 14.00 2.05 a 7.67 1.29 b 9.42 1.37 b 3.91 0.0152 Pan 0.19 0.10 b 0.56 0.16 ab 1.31 0.37 a 1.13 0.39 a 3.27 0.0275 Pitfall 1.00 0.39 2.50 0.44 2.08 0.82 2.83 0.60 1.75 0.1720 Araneae Sticky 1.08 0.38 a 0.17 0.11 b 0.17 0.11 b 0.25 0.013 b 4.14 0.0119 Pan 0.44 0.16 a 0.19 0.10 ab 0.06 0.06 b 0.13 0.09 b 2.26 0.0909 Pitfall 0.33 0.14 0.75 0.21 1.08 0.29 1.00 0.30 1.81 0.1613 Total natural enemies 3 Sticky 11.08 1.48 b 18.58 2.07 a 10.25 1.63 b 12.08 2.23 b 3.88 0.0157 Pan 0.94 0.23 1.75 0.46 1.81 0.42 1.60 0.47 0.97 0.4154 Pitfall 1.83 0.65 b 4.25 0.65 ab 3.42 1.03 ab 5.17 0.90 a 2.80 0.0520 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P according to LSD test. No letters in rows indicate no differences among means. 1 Analysis of variance; F and P values; Sticky cards (d. f. = 3, 41); Pan traps (d. f. = 3, 57) ; Pitfall traps (d. f. = 3, 41). 2 Coccinellidae, Staphylinidae, Carabidae, Cantharidae, and Mordellidae. Predatory Coleoptera were rarely observed in pan traps, and were included in Total natural enemies. 3 Aranae, Formicidae, Mutillidae, microhymenoptera, macrohymenoptera, predatory Coleoptera, Diptera (Asilidae, Bombyliidae, Dolichopodidae and Syrphidae), and Hemiptera ( Geocoris spp., Anthrocoridae, and Reduviidae).
69 Table 35 Selected arthropod groups (number per trap) recovered on sticky card s, pan traps and pitfall traps within border treatments, 2008. Treatments ANOVA 1 Arthropod group Trap Bare ground Pigeon pea Sorghum Weeds F P Orthoptera Sticky 0.17 0.17 0.33 0.26 0.00 0.00 0.58 0.26 1.46 0.2393 Pan 0.06 0.06 0.25 0.11 0.19 0.10 0.06 0.06 1.12 0.3476 Pitfall 1.08 0.31 1.92 0.63 1.00 0.46 1.25 0.39 0.91 0.4428 Hemiptera Aleyrodidae Sticky 5.00 1.29 a 3.25 0.93 ab 1.08 0.68 b 2.41 0.65 b 3.42 0.0258 Aphididae Sticky 11.58 2.21 9.92 2.97 6.25 1.80 7.83 2.65 0.87 0.4631 Pan 0.44 0.22 b 0.88 0.29 ab 1.56 0.34 a 0.77 0.20 b 3.01 0.0374 Pitfall 1.17 0.60 2.08 0.81 0.67 0.36 3.50 1.71 1.46 0.2402 Auchenorrhyncha 2 Sticky 3.25 0.91 b 7.92 1.16 a 2.91 0.83 b 7.83 1.85 a 4.87 0.0055 Pan 0.06 0.06 b 1.06 0.40 a 0.44 0.22 ab 1.31 0.47 a 2.89 0.0431 Pitfall 1.50 0.44 20.25 16.32 1.00 0.33 9.08 4.61 1.12 0.3536 Cicadellidae Sticky 1.75 0.52 b 3.00 0.82 b 2.92 0.83 b 6.83 1.95 a 3.61 0.0211 Membracidae Sticky 1.42 0.81 b 4.50 1.13 a 0.00 0.00 b 0.67 0.31 b 7.45 0.0004 Heteroptera 3 Sticky 0.17 0.17 0.33 0.26 0.00 0.00 0.58 0.26 1.46 0.2393 Pan 0.13 0.09 0.06 0.06 0.06 0.06 0.00 0.00 0.73 0.5379 Pitfall 0.00 0.00 0.00 0.00 0.00 0.00 0.17 0.11 1.14 0.2455 Coleoptera Altica spp. Sticky 0.67 0.28 a 0.08 0.08 b 0.00 0.00 b 0.00 0.00 b 4.62 0.0071 Pan 0.06 0.06 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.3995 Non predatory 4 Sticky 0.25 0.13 0.08 0.08 0.42 0.19 0.08 0.08 1.46 0.2397 Pan 0.13 0.09 0.00 0.00 0.13 0.09 0.06 0.06 0.81 0.4918 Pitfall 0.00 0.00 0.25 0.13 0.33 0.26 0.25 0.18 1.17 0.3390 Lepidoptera Sticky 0.08 0.08 0.17 0.11 0.00 0.00 0.42 0.19 2.13 0.1117 Pan 0.13 0.09 0.13 0.09 0.13 0.09 0.29 0.15 0.64 0.5949 Pitfall 0.08 0.08 0.17 0.11 0.00 0.00 0.00 0.00 1.33 0.2776
70 Table 3 5. Continued Diptera Microdiptera Sticky 9.42 1.31 b 14.92 2.98 a 5.33 1.26 b 9.75 1.10 b 4.83 0.0057 Pan 1.00 0.35 1.94 0.50 2.31 0.62 2.90 0.73 1.91 0.1381 Pitfall 1.41 0.26 b 4.92 1.50 a 2.67 0.64 ab 3.17 0.92 ab 2.31 0.0900 Thysanoptera 5 Sticky 2.08 0.74 ab 1.75 0.71 ab 0.25 0.18 b 2.92 0.87 a 2.56 0.0678 Pan 0.25 0.11 0.19 0.10 0.00 0.00 0.23 0.14 1.36 0.2651 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P ding to LSD test. No letters in rows indicate no differences among means. 1 Analysis of variance; F and P values; Sticky cards (d. f. = 3, 41); Pan traps (d. f. = 3, 57); Pitfall traps (d. f. = 3, 41). 2 Cercopidae, Cicadellidae, Delphacidae, and Membracidae. 3 Herbivorous Aleyidae, Lygaeidae, Miridae, Pentatomidae, Pyrrochoridae, and Rhopalidae. 4 Chrysomelidae, Elateridae, Tenebrionidae 5 Primarily Thripidae that were rarely observed in pitfall traps.
71 Table 36. Selected beneficial arthropod groups (number per trap) recovered on sticky card s, pan traps and pitfall traps within squ ash crop, 2007. Treatments ANOVA 1 Arthropod group Trap Bare ground Pigeon pea Sorghum Weeds F P Coleoptera Preda tory 2 Sticky 1.33 0.44 0.75 0.33 1.17 0.35 0.75 0.22 0.84 0.4782 Pitfall 1.17 0.46 1.00 0.30 1.75 0.71 0.42 0.29 0.59 0.6272 Diptera Dolic hopodidae Sticky 4.33 0.92 3.50 0.75 3.00 0.83 3.25 0.61 0.56 0.6413 Pan 0.25 0.18 0.00 0.00 0.00 0.00 0.00 0.00 1.91 0.1427 Pitf all 3.25 1.05 2.67 0.48 1.83 0.71 2.17 0.59 0.74 0.5332 Hymenoptera Formicidae Sticky 0.33 0.14 0.33 0.19 0.00 0.00 0.33 0.19 1.26 0.3002 Pan 0.00 0.00 0.00 0.00 0.08 0.08 0.00 0.00 1.00 0.4402 Pitfall 3.37 1.55 5.92 1.26 4.50 1.48 5.31 1.35 0.31 0.8161 Microhymenoptera Sticky 23.33 4.47 17.50 2.70 15.58 2.64 17.50 2.64 1.12 0.3522 Pan 0.08 0.08 0.33 0.19 0.42 0.19 0.17 0.11 1.00 0.4004 Pitfall 2.25 0.64 2.08 0.68 1.08 0.29 3.58 1.00 2.17 0.1066 Araneae Sticky 0.83 0.27 b 1.00 0.41 b 2.08 0.45 a 1.00 0.30 b 2.34 0.0873 Pan 0.25 0.13 ab 0.08 0.08 b 0.42 0.15 a 0.00 0.00 b 2.79 0.0525 Pitfall 0.83 0.27 1.00 0.44 0.67 0.19 0.67 0.23 0.32 0.8085 Total natural enemies 3 Sticky 31.17 4.73 24.00 3.51 22.42 3.47 23.42 3.09 1.17 0.3326 Pan 0.67 0.23 0.42 0.26 0.92 0.26 0.17 0.11 2.03 0.1247 Pitfall 8.25 1.46 7.67 0.98 6.75 1.01 7.33 1.19 0.31 0.8177 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P according to LSD test. No letters in rows indicate no differences among means. 1 Analysis of variance; F and P values; Sticky cards (d. f. = 3, 41); Pan traps (d. f. = 3, 41); Pitfall traps (d. f. = 3, 41). 2 Coccinellidae, Staphylinidae, Carabidae, Cantharidae, and Mordellidae. Predatory Coleoptera were rarely observed in pan trap s, and were included in Total natural enemies. 3 Aranae, Formicidae, Mutillidae, microhymenoptera, macrohymenoptera, predatory Coleoptera, Diptera (Asilidae, Bombyliidae, Dol ichopodidae and Syrphidae), and Hemiptera ( Geocoris spp., Anthrocoridae, and Reduviidae).
72 Table 37. Selected ar thropod groups (number per trap) recovered on sticky card s, pan traps and pitfall traps recovered in adjacent squash crop, 2007. Treatments ANOVA 1 Arthropod group Trap Bare ground Pigeon pea Sorghum Native Weeds F P Orthoptera Sticky 0.42 0.19 0.83 0.44 0.08 0.08 0.42 0.19 1.30 0.2872 Pan 0.00 0.00 0.00 0.00 0.00 0.00 0.08 0.08 1.00 0.4025 Pitfall 1.08 0.34 1.17 0.51 1.42 0.31 1.83 0.41 0.72 0.5471 Hemiptera Aleyrodidae Sticky 14.75 3.39 12.08 3.16 8.33 1.75 9.92 1.60 1.07 0.3712 Aphididae Sticky 21.25 5.11 21.25 5.54 18.75 4.42 14.83 4.12 0.41 0.7490 Pan 0.17 0.11 0.25 0.18 1.08 0.66 0.25 0.13 1.61 0.2011 Auchenorrhyncha 2 Sticky 6.50 0.77 ab 7.42 1.28 ab 4.17 0.76 b 8.25 1.57 a 2.39 0.0828 Pan 0.42 0.26 0.25 0.13 0.75 0.18 0.67 0.28 1.08 0.3688 Pitfall 1.58 0.48 1.50 0.58 1.58 0.36 1.83 0.63 0.06 0.9818 Cicadellidae Sticky 5.58 0.69 6.17 1.13 3.83 0.74 6.92 1.50 1.50 0.2291 Membracidae Sticky 0.75 0.22 1.08 0.29 0.25 0.13 0.75 0.33 1.97 0.1328 Heteroptera 3 Sticky 0.42 0.15 0.50 0.26 0.25 0.13 0.17 0.11 0.79 0.5040 Pan 0.17 0.11 0.05 0.26 0.17 0.11 0.00 0.00 1.82 0.1594 Pitfall 2.58 0.76 1.67 0.63 1.17 0.37 1.92 0.69 0.82 0.4907 Coleoptera Altica spp. Sticky 3.00 1.27 2.83 0.43 4.33 1.81 3.08 1.29 0.23 0.8716 Pan 0.08 0.08 ab 0.25 0.13 a 0.00 0.00 b 0.00 0.00 b 2.41 0.0863 Non predatory 4 Sticky 0.75 0.37 1.08 0.44 1.00 0.30 0.50 0.20 0.58 0.6289 Pan 0.25 0.13 0.25 0.13 0.00 0.00 0.25 0.18 0.92 0.4396 Pitfall 1.00 0.46 0.67 0.28 1.08 0.50 0.42 0.29 0.59 0.6272 Lepidoptera Sticky 0.08 0.08 0.42 0.23 0.67 41 0.50 0.15 0.94 0.4301 Pan 0.42 0.19 0.33 0.17 0.17 0.11 0.08 0.08 0.98 0.4133 Pitfall 0.08 0.08 0.42 0.23 0.25 0.18 0.17 0.11 0.80 0.5033
73 Table 3 7. Continued Diptera Microdiptera Sticky 15.08 4.82 8.42 2.04 15.08 3.40 11.92 2.77 0.84 0.4819 Pan 1.42 0.56 ab 0.42 0.26 b 1.17 0.34 ab 2.17 0.61 a 2.07 0.0734 Pitfall 2.08 1.47 1.92 0.63 1.17 0.34 1.25 0.41 0.30 0.8264 Thysanoptera 5 Sticky 10.08 1.92 5.42 1.56 4.75 1.26 15.67 7.62 1.50 0.2280 Pan 0.08 0.08 0.08 0.08 0.17 0.17 0.00 0.00 0.44 0.7250 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P ding to LSD test. No letters in rows indicate no differences among means. 1 Analysis of variance; F and P values; Sticky cards (d. f. = 3, 41); Pan traps (d. f. = 3, 41); Pitfall traps (d. f. = 3, 41). 2 Cercopidae, Cicadellidae, Delphacidae, and Membracidae. 3 Herbivorous Aleyidae, Lygaeidae, Miridae, Pentatomidae, Pyrrochoridae, and Rhopalidae. 4 Chrysomelidae, Elateridae, Tenebrionidae 5 Primarily Thripidae that were rarely observed in pitfall traps.
74 Table 38. Selected beneficial arthropod groups (number per trap) recovered on sticky card s, pan traps and pitfall traps within squash crop, 2008. Treatmen ts ANOVA 1 Arthropod group Trap Bare ground Pigeon pea Sorghum Native Weeds F P Coleoptera Predatory 2 Sticky 0.25 0.13 b 0.17 0.11 b 1.25 0.55 a 0.42 0.19 ab 2.74 0.0555 Pitfall 0.08 0.08 0.00 0.00 0.17 0.11 0.33 0.142 2.05 0.1214 Diptera Dolichopodidae Sticky 1.25 0.46 1.25 0.55 0.92 0.26 2.33 0.61 1.53 0.2202 Pan 0.06 0.06 0.06 0.06 0.00 0.00 0.00 0.00 0.66 0.5830 Pitfall 0.17 0.11 0.42 0.19 0.25 0.13 0.42 0.19 0.59 0.6248 Hymenoptera Formicidae Sticky 0.17 0.11 ab 0.00 0.00 b 0.42 0.19 a 0.08 0.08 ab 2.22 0.0997 Pan 0.13 0.09 0.13 0.09 0.15 0.09 0.00 0.00 0.81 0.4942 Pitfall 12.75 2.88 6.08 1.37 7.75 1.56 10.08 2.78 1.66 0.1915 Microhymenoptera Sticky 9.33 1.61 8.25 1.11 10.42 2.36 7.08 1.18 0.73 0.5423 Pan 0.44 0.18 0.75 0.25 0.54 0.20 0.56 0.22 0.36 0.7822 Pitfall 2.00 0.49 2.08 0.58 1.80 0.52 2.92 0.62 0.73 0.5387 Araneae Sticky 1.08 0.29 0.67 0.28 0.50 0.23 0.33 0.14 1.69 0.1845 Pan 0.31 0.15 0.25 0.14 0.35 0.12 0.19 0.10 0.31 0.8186 Pitfall 0.25 0.13 0.33 0.14 0.17 0.11 0.17 0.11 0.41 0.7485 Total natural enemies 3 Sticky 12.08 2.06 10.50 1.47 13.83 2.44 10.25 1.42 0.74 0.5331 Pan 0.94 0.21 1.31 0.30 0.98 0.25 0.81 0.25 0.68 0.5689 Pitfall 2.58 0.54 3.17 0.81 2.50 0.68 4.08 0.88 0.96 0.4193 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P according to LSD test. No letters in rows indicate no differences among means. 1 Analysis of variance; F and P values; Sticky cards (d. f. = 3, 41); Pan traps (d. f. = 3, 57); Pitfall traps (d. f. = 3, 41). 2 Coccinellidae, Staphylinidae, Carabidae, Cantharidae, and Mordellidae. Predatory Coleoptera were rarely observed in pan trap s, and were included in Total natural enemies. 3 Aranae, Formicidae, Mutillidae, microhymenoptera, macrohymenoptera, predatory Coleoptera, Diptera (Asilidae, Bombyliidae, Dol ichopodidae and Syrphidae), and Hemiptera ( Geocoris spp., Anthrocoridae, and Reduviidae).
75 Table 39. Effect of treatments on s elected arthropod groups (number per trap) recovered on sticky card s, pan traps and pitfall traps recovered in adjacent squash crop, 2008. Treatments ANOVA 1 Arthropod group Trap Bare ground Pigeon pea Sorghum Native Weeds F P Orthoptera Sticky 0.08 0.08 0.08 0.08 0.25 0.18 0.00 0.00 0.92 0.4377 Pan 0.00 0.00 0.06 0.06 0.08 0.07 0.13 0.09 0.70 0.5586 Pitfall 0.17 0.11 0.58 0.29 0.92 0.34 0.58 0.19 1.71 0.1805 Hemiptera Aleyrodidae Sticky 13.00 3.17 15.50 4.53 13.92 4.17 8.83 2.56 0.61 0.6147 Aphididae Sticky 8.67 2.21 8.83 1.36 10.67 1.74 8.25 1.39 0.39 0.7595 Pan 0.38 0.15 0.69 0.24 0.92 0.28 0.94 0.21 1.31 0.2790 Auchenorrhyncha 2 Sticky 2.33 0.63 3.08 0.60 2.33 0.88 4.42 0.91 1.84 0.1558 Pan 0.13 0.09 0.25 0.11 0.00 0.00 0.19 0.10 1.51 0.2222 Pitfall 1.42 0.66 b 5.08 2.51 ab 0.75 0.22 b 9.58 0.42 a 2.68 0.0595 Cicadellidae Sticky 1.50 0.40 b 1.17 0.35 b 1.92 0.82 ab 3.42 0.83 a 2.58 0.0665 Membracidae Sticky 0.83 0.35 ab 1.58 0.50 a 0.25 0.13 b 0.58 0.29 b 2.72 0.0571 Heteroptera 3 Sticky 0.17 0.11 0.25 0.13 0.08 0.08 0.08 0.08 0.62 0.6037 Pan 0.06 0.06 0.00 0.00 0.06 0.06 0.00 0.00 0.70 0.5537 Pitfall 2.58 0.76 1.67 0.63 1.17 0.37 1.92 0.69 0.82 0.4907 Coleoptera Altica spp. Sticky 0.83 0.42 0.33 0.19 0.83 0.37 0.67 0.22 0.54 0.6603 Pan 0.06 0.06 0.00 0.00 0.06 0.06 0.13 0.09 0.70 0.5537 Non predatory 4 Sticky 0.33 0.22 0.00 0.00 0.17 0.11 0.08 0.08 1.15 0.3416 Pan 0.06 0.06 0.13 0.09 0.06 0.06 0.13 0.09 0.23 0.8719 Pitfall 1.75 0.79 1.58 0.61 0.58 0.34 2.08 0.98 0.76 0.5221 Lepidoptera Sticky 0.08 0.08 0.08 0.08 0.08 0.08 0.00 0.00 0.39 0.7605 Pan 0.13 0.09 0.06 0.06 0.06 0.06 0.00 0.00 1.46 0.2078 Pitfall 0.00 0.00 0.00 0.00 0.08 0.08 0.17 0.11 1.33 0.2776
76 Table 3 9. Continued Diptera Microdiptera Sticky 8.50 1.04 9.00 0.65 9.33 0.69 7.83 1.10 0.52 0.6692 Pan 0.56 0.20 0.75 0.27 0.85 0.22 1.13 0.29 0.89 0.4538 Pitfall 0.83 0.30 1.08 0.31 1.33 0.40 2.00 1.13 0.63 0.5990 Thysanoptera 5 Sticky 0.14 0.16 1.58 0.47 1.25 0.49 1.42 0.71 0.10 0.9605 Pan 0.63 0.20 0.31 0.15 0.44 0.16 0.19 0.10 1.40 0.3804 Data are means standard errors of 4 replications. Means in rows followed by the same letter do not differ (P ding to LSD test. No letters in rows indicate no differences among means. 1 Analysis of variance; F and P values; Sticky cards (d. f. = 3, 41); Pan traps (d. f. = 3, 57); Pitfall traps (d. f. = 3, 41). 2 Cercopidae, Cicadellidae, Delphacidae, and Membracidae. 3 Herbivorous Aleyidae, Lygaeidae, Miridae, Pentatomidae, Pyrrochoridae, and Rhopalidae. 4 Chrysomelidae, Elateridae, Tenebrionidae 5 Primarily Thripidae that were rarely observed in pitfall traps.
77 CHAPTER 4 PEST MANAGEMENT STRATEGIES AND LIVELIHOOD ACTIVITIES OF SMALL SCALE AGRICULTURALISTS IN HAITIS NORTHERN CENTRAL PLATEAU Introduction Once regarded by colonialists as the jewel of the Caribbean, the Republic of Haiti ranks as the poorest country in the Western Hemisphere (Jadotte 2006). Widespread poverty in rural Haiti is exacerbated by the continuous decline within the agricultural sector, severe environmental degradation, prolonged political and economic instability, and a dearth of employment opportunities ( Smucker et al. 2002). An estimated 9 million people live in Haiti and more than twothirds of the population is unemployed or underemployed (CIA 2009). Despite billions of dollars in foreign aid (Buss and Gardner 2006) p overty rates in the Central and Western Haiti were an estimated at 77% in 1996 (Wiens and Sobrado 1998)1 and more than 54% of Haitians live in abject poverty (CIA 2009). Since 1998, the value of Haitis currency, the Haitian Gour de, has dropped more than twofold2. Agriculture is the predominant economic activity in rural Haiti ( Smucker et al. 2002, Chen and Murray 1976) When agriculture fails, natural disaster strikes, or capricious market prices make nutritional staples like rice and beans unattainable, how does a rural Haitian household survive? Through conversations with farmers, using a methodology called a Sondeo (Hildebrand 1981) we sought to understand the Disclaimer: All information presented, is based on perceptions of participating agriculturalists and may not necessarily reflect objective realities. Any mention of organizations or individuals involved in agricultural and community development are not meant as an endorsement, nor is the intention of this paper to criticize the merits of their efforts. 1 These data were based on the local costs associated with obtaining the FAO minimum daily nutritional standard. 2 The exchange rate on August 3, 1998, was 16.046 Haitian Goudes (HTG) = $1.00 US Dollar (USD). More than ten years later, on March 5, 2009, 40.495 HTG = $1.00 USD.
78 socioeconomic and agricultural framework that enables households to endure these challenges year after year. Rationale Billions of dollars have been invested in development efforts in Hait i over the past three decades ( Buss and Gardner 2006), yet poverty persists. One does not have to look very far to find an abandoned aid project, half built school or well intentioned, but poorly realized attempt to bring assistance and relief to rural Haitians. Perhaps one of the reasons so many development projects are unsuccessful in Haiti is that aid groups have failed to understand the livelihood system s of the people they are seeking to assist. Understanding a livelihood system is important for identifying and defining problems ; however it is even more critical for shaping solutions ( Collinson 2001) The goal of this study was to develop a broad understanding of the household livelihood system of the greater Bohoc region by identifying and descr ibing key incomegenerating or provisionary activities employed by households and by recognizing major socioeconomic constraints that limit production and govern decision making. Because agricultural production is such a key component of household liveli hood system s in rural Haiti (Chen and Murray 1976) the main objectives of this study were to make an appraisal of the local cropping system, summarize major factors that limit production, and then identify methods implemented by farmers to meet these chal lenges. One of the areas of focus for this study was pest management. Although the use of highly toxic pesticides has been documented in southern Haiti (Bishop 1995) and in neighboring Dominican Republic (Murray 1994), pesticide use in north central Hait i has not been recorded. If farmers in this region employed agricultural chemicals, we hoped to understand their experience with pesticide use including their access to
79 synthetic and natural products, training, and personal protection equipment (PPE), as well as the investment (costs and labor implications) associated with their usage. We discuss farmer opinions regarding pesticide efficacy and safety and document specific product use. This study will be presented to local nongovernmental organizations (NGOs )3 and missions working in this region of Haiti as a summary of the challenges, needs, values, and activities of households in the community where they work. It will also provide an assessment of agricultural activities, training gaps, and areas for potential program enhancement or agricultural research. The baseline data gathered during this study will be integrated into a livelihood systems analysis project conducted by some of the authors during 2010. Economic data, including livestock and other household livelihood strategies gathered during the S ondeo are not included in this document. Methodology A Sondeo (Hildebrand 1981) is a participatory rural appraisal method that uses discussions rather than formal questionnaires to obtain informati on about topics of interest (Breuer et al. 2008). Interviewers are grouped into multi disciplinary teams that refrain from taking notes or referring to a list of questions so that interviews can be truly conversational (Breuer et al. 2008). The role of the Sondeo team in this qualitative research is to listen, observe, and direct conversations to meet a set of objectives while allowing the interviewee the opportunity to discuss issues of great concern to him or her. The Sondeo approach does not necessaril y yield quantifiable data; however, it is 3 The abbreviation ONG ( Organisation NonGouvernemental ) is used by the Haitian government.
80 an effective, inexpensive, and rapid assessment tool that provides information that is rarely obtained from traditional surveys (Hildebrand 1981). In this case, the Sondeo team consisted of precollegiate Haitian nationals who served as guides and translators, as well as graduate students and professionals from a number of backgrounds including agriculture, animal science, anthropology, economics, education, entomology, forestry, law, and medicine4. A total of 12 members of the Sondeo team were organized into three multidisciplinary discussion groups with 3 4 members each. I nterviewers rotated each morning and afternoon so that every member had the opportunity to learn from the other team members Informal conversations with farmers and the members of their households were conducted over a 4day period. The Sondeo team identified 5 key agricultural environments, and each day communities within those environments were selected for conducting conversational intervi ews. Households and individual farmers were randomly selected by the teams as they walked along corridors within the community. Individual conversations lasted between 15 minutes and 1.5 hours, and usually took place in the yard ( lakou) Many farmers ga ve impromptu tours of the crops growing in their lakou or fields ( jaden) if they were nearby. However, the main gardens were usually some distance from the house. Each evening, team members met to discuss the conversational interviews conducted earlier in the day. Through these discussions, the team identified general trends, informational gaps, and cultural nuances from which new questions emerged for the next day of interviews. A total of 7 communities were visited within a 19 km2 area 4 Members of the Sondeo team included: H. N. HansPetersen, R. Slaughter, G. VanSchoyck, B. A. VanSchoyk, C. A. Curilla, C olin Wilson, C arla Wilson, E. P. Campbell, J. T. HansPetersen, D. Saint Fleure, W. Des Auguste, and D. Mompremier.
81 surrounding the community of Bohoc The teams conversed with a total of 38 farmers/farmer families, two agricultural exten sion agents, and various vendors at the Bohoc market Results Study Area The community of Bohoc lies roughly 350 meters above sea level and is transected by National Route 3 and the Bohoc River which marks the boundary between two of Haitis ten administrative departments, the Central Plateau (Plato Santral) and Department of the North (Depatman N)5. T he Bouy aha River served as the western limit of the study area, while the Masif du N mountains marked the eastern border Bohoc is located 22 km north of Hinche the capital of the Central Plateau, on National Route 3, and 13 km south of Pignon, a city equipped with a small air strip, hospital facilities, a restaurant, and banking services. Agriculture is the predominant economic activity in the greater Bohoc region followed by commerce, which is primarily the work of women. Residents of this area were likely to work land in at least two of five environments: hillsides, irrigated lowland, flatland, savann, and the lakou surrounding their home. Many of the people encountered had multiple pieces of land that they either owned or sharecropped in addition to the land where their house was located. This is consistent with findings by other researchers who agree that most rural Haitian farmers are the owner operators of small, fragmented landholdings of poor soil fertility ( Smucker et al. 2002). However, it was not always clear to our team whether the land was owned or rented because some 5 Residents living on the north side of the Bohoc River report to GrandeRivire duNord an arrondissment in the Department of the North, w hile those living on the south side of the river must go to Hinche.
82 individuals were reluctant to share that information or made exaggerated statements about their landholdings. In rural Haiti, land area is measured in units called kawo6, equivalent to approximately 1.29 hectares or roughly 3.19 acres ( FAO 2007). Wiens and Sobrado ( 1998) reported that 75 % of rural farmers in Haiti have access to land, but suggest ed that the average farm size was less than 2 hectares di stributed over about 3 s maller parcels called gardens ( jaden) when under cultivation. The quality of soil on the average farm varies, and usually only a fraction of the land is of good to mixed quality ( Smucker et al. 2002; Wiens and Sobrado 1998). The majority of the gardens and all of the lakou have clear boundaries that are delineated by a cactus like living fence called rakt (Euphorbia sp ) interplanted with tree species. This was especially true in the irrigated regions where the soil was generally more fertile and where gardens were interspersed with housing or were along major footpaths. Climate and Environment Central Haiti has bimodal rainy seasons (Butterman 1997). The early rainfall season begins in May and extends until June. A mid summer drought typically during the month of July, especially during El Nio years, is followed by a late rainfall season that spans August to November ( Butterman 1997). Few precipitation events occur in the months of December and January. Because of its location in the Caribbean basin, Haiti is subject to hurricanes between June and November, the key growing season for rainfed farms Farm ers indicated that the seasons have been changing since the mid1970s They used to have enough rain to grow black beans ( pwa nwa) and lima beans 6 In French, the kawo unit is called a carreau, and plural units are carreaux de terre
83 ( pwa boukousou) They reported that periods of drought have become more frequent and severe and that the rai ny seasons are generally shorter. Hurricanes and drought have killed livestock and decimated three consecutive harvests in this region. Steep, denuded mountains bordering the study area were at one time covered with trees. During several conversations, interviewees mentioned that pine trees ( Pinus spp.) could no longer be found in the area. Older farmers identified deforestation as the cause of soil degradation associated with yield declines. Others mentioned the ability of trees to hold the soil, and noted that they are losing their soil in the absence of trees Erosion mitigation berms were not observed in the gardens visited during the investigation. H owever, a series of stone erosion retention walls on the hills above Lapil a was being constructed and financed by a NG O Cropping S ystem Agriculture in this region is characterized by extensive polyculture. The cropping pattern and placement of specific crops reflects the acute awareness of topographical fertility and micoclimate features of a given parcel of land ( Smucker et al. 2002). Crop diversity within a traditional mixed cropping system is a risk management strategy employed by subsistence farmers that expands both the duration of harvest and the availability of food resources for the household ( Smucker et al. 2002). Root crops like cassava, cocoyam, taro, and yam can be stored below ground and harvested year round. Tree species that provide poles, construction materials, charcoal, or fruit are integrated into most of these agricultural landscapes; however, the bulk of the trees were found in the lakou where they can be protected from machetes, loose animals, and people making charcoal. Over 20 fruit and nut trees found in this area complement the local diet (Table 41 ). The harvests of som e of these fruits occur during the leanest
84 months of the year. Mangoes are a particularly important fruit during the month of July, which was identified by several farmers as the season when crops are in the ground but not ready to harvest There has been an emphasis by local NGOs and nurseries on grafting superior varieties or cultivars that extend the harvest season of mangoes, citrus, and avocado. Some farmers have had their full grown mango trees topgrafted with improved varieties as well. Fruit and nut harvests are for household consumption first and then they are shared with friends and neighbors or sold. Staples of the local diet include cereal crops such as maize, millet legumes including peanuts and pigeon peas and bean cultivars, as well as various root crops and imported rice7 ( Table 42 ). During November of 2008 we observed mixed cropping systems including any combination of late season millet, pigeon pea, sweetpotatoes, peanuts, and cassava planted simultaneously in an orderly but not linear fashion. Maize and millet were both being harvested and prepared for storage. A typical rainfed field might have been planted with one marmit8 of maize, two godes millet and one gode of pigeon pea. A few farmers intercrop maize and cowpea ( Vigna unigulata) successfully, which the extension agent believes could help area farmers. In general, the food grown by these farmers is for their familys consumption. Sugarcane ( Saccharum officinarum) is the main cash crop of t he central plateau. Local distilleries purchase cane syrup by the full drum for $500 HD per 55 gal. (<$63.00 USD). Every farmer working land that would support sugarcane grew some quantity to 7 During the 1960s this area produced dryland rice for local consumption. This has diminished so much that only one farmer we encountered grew rice. 8 Marmit is the basic volumetric unit of measurement in the Haitian market system based upon the old sta ndard no.10 can of roughly 3 quarts. Seven godes, which are essentially an enamel coffee mug, fill one marmit
85 ensure access to currency and a quick source of calories, especially during leaner months. Sugarcane also provides useful by products including bagasse9, which can be fed to cattle, burned during syrup processing, or composted (however, only one of the farmers we met with described this latter use). Cane leaves ar e left on the surface of the soil after the harvest to suppress weeds, mitigate erosion, and retain soil moisture. Because sugarcane is a perennial, the same planting can be harvested through multiple seasons; however, one farmer reported that the number of drums of syrup significantly declines with each harvest. Many farmers practice a fallow to restore some of the fertility to the land after about three sugarcane harvests. Sugarcane culture is labor intensive and is the regions only consistently monoc ultured cropping system. Farmers will often clear their fields using a controlled burn prior to planting. The practice of burning also helps provide control of a plant disease referred to as chabon10 which turns newly emerged shoots black. Depending on the size of the land and the resources available, a farm family may pay someone to help with cutting and processing the juice. These steps are discussed more thoroughly in the economics section of this paper. Several sugarcane presses were ac tively processing cane during November 2008. Agricultural Landscape Farmers in the greater Bohoc area typically own or sharecrop multiple pieces of land in different terrains that they use for different purposes including raising crops, animal grazing, fallow, harvesting forestry products and charcoal production. It is common for men and women to travel at least an hour to their fields each way. Five 9 Bagasse is a byproduct of the canemilling process. Once the juice has been pressed out of the stalk, bagasse is left. It can be used as fuel to boil the sugarcane juice down to a syrup, composted, or fed to livestock. 10 Chabo n translates as charcoal in English. The same word is used for charcoal and anthrax.
86 agricultural environments were identified during the Sondeo: flatland, hillsides, irrigated lowland, sava nn and the lakou. Flatland. What we have identified as the flatland area includes the communities of LaJuene, Lapila, Terre Glisse and Bohoc, which border National Route 3 and were the most densely populated communities visited. Because there were more homes packed closely together, and the home is the site where the majority of trees are planted, this region generally has much greater tree cover than some of the other areas visited. A network of f ootpaths lined with Euphorbia spp. and a variety of woody plants cover the entire zone. The soil of the flatland area is characterized by the alluvial deposits once held by trees and vegetation on the mountainsides and appears to be generally darker, less rocky, and of greater fertility than the soil of many of the areas we visited. The flatter topography of this environment likely slows soil degrading erosion. In general, n o irrigation is available in this area, but we observed at least one spring that was providing water to a sugarcane field in Lapila. Sugarcane was the primary crop grown in this area, and several cane mills were operating during our visit. Farmers in this area typically pay to have their land plowed by oxen rather than prepare the land by hand. In addition to sugarcane, mixed cropping sys tems, including various combinations of maize, millet, pigeon pea, sweet potato, peanuts, and cassava were observed. Hillsides The community of Skadi extends through a ravine that connects to the town of Laviktwa and terminates just upstream from where the Bohoc River crosses National Route 3 near the Bohoc Market. The steep hillsides defining the valley have suffered severe erosion. Several caves can be found in the rocky crags along the sides of the ravine. Ol der farmers in the area recalled these hillsides having much more soil
87 and much larger harvests Maize and millet are planted among the rocky outcroppings of limestone bedrock along the steep sides of the ravine. These sites are prepared by hand using a hoe, simply because the terrain is too treacherous and rocky for the oxen and plow. Sugarcane is not planted along the sloping hillsides for the same reason. However, several hectares of what are considered tillable plateaus are present on the hills of t he south side of the ravine. In Skadi farmers grew a selection of rainfed crops including short and long season millet, maize, cassava, pigeon pea, and sweetpotato. The diversity and quality (size and overall health) of crops as well as apparent soil fertility and quality progressively improved down the side of the ravine. Midway down the mountain, the soil was rocky, but no longer characterized by outcroppings of bedrock. The land at the base of the ravine on either side of the stream ( Bohoc River) is highly valued due to the soil quality and the close proximity to the water. Here vegetables, beans, tubers, and a limited amount of sugarcane were grown in addition to the other crops. However, the majority of the agricultural land in Skadi was locat ed along the sides of the ravine and the plateaus far above the fertile river basin. Farmers in this area did not have improved irrigation and desired assistance with generator operated pumps and watering cans. Farmers in the flatland area (Terre Glisse) reflected on the poor soil quality and slopes that limit the ability of farmers in Skadi to produce sugarcane. They stated that farmers in Skadi, who cannot grow sugarcane, are envious of the land in Terre Glisse. Irrigated lowland. The irrigated regi on of the community of L t b B ohoc, literally the other side of Bohoc, is located in the alluvial flood plain of the Bohoc River. This lowlying area has dark rich soil and is the breadbasket of the community. In 2004 World Vision engineers designed an open irrigation system about 1.2 km in length that
88 allows food to be grown in this area year round in an intensive polyculture system that included vegetables, pulse, grain, and tubers crops (Table 42, 43) The most valuable vegetable according to these growers is cabbage ( chou) which is the only crop to which they can afford to apply commercial pesticides and fertilizer. They begin planting cabbage in August and September and continue planting through November. Seedlings are grown in a nursery bed w ithin the field and then transplanted by hand into a hole fi l led with bat guano. After the first harvest, they allow the plants to resprout, which produces multiple smaller heads that still can be sold in the market. Black beans are planted in this area following the second cabbage harvest, but must be planted prior to January according to one farmer ; otherwise the beans will not form pods. S imultaneous intercropping of cabbage, okra, and tomatoes; cabbage, tomatoes, and cassava; and cabbage and peanuts was observed in gardens in Lt bBohoc. Additionally, plantains, bananas, coconut trees, taro and papaya were planted in the borders of the gardens and along the irrigation canals. This is another example of what Smucker et al. (2000) describes as Hait ian farmers acute awareness of microclimate variation within their limited pieces of land. Because food can be grown year round with irrigation, farmers practice a crop rotation program that includes black beans, onions, and maize, in addition to the ca bbage, tomato, okra, and peanuts already in the ground. In one plot, r ows of cassava a perennial crop, were interplanted with short term vegetable crops so that they could plow the annual vegetable rows post harvest and r eplant without disturbing the row s of manioc. This was the only area where farmers discussed their s easonal crop rotation plans, soil improvement and where crops were planted in rows.
89 Savann. The savann is the largest and most remote region visited during the Sondeo. Access to National Route 3 takes at least 1.5 hours on foot from the savann villages of Monak and Don. Houses and gardens were more dispersed in this area. Many individuals who live in mo re highly populated communities like Bohoc and LaJeune own land in this region for grazing their animals, which contributes considerably to erosion in this area. Tree cover is sparse except in areas where there are homes. An invasive nonnative grass species called Madanm Michl ( Themeda quadrivalvis) which is unpalatable to livestock, covers much of the undeveloped and fallow land. The open plains, drought, and dry grass often result in sweeping fires that burn across the savann, consuming homes, crops, and livestock. Denuded, overgrazed rolling hills are rife with ravines excised by water from major weather events. Although, the ravines provide a fertile and moist microclimate that is readily exploited by planting with sugarcane, taro, bananas, and plantains, these areas are most at risk during a weather event. Several farmers described losses from mudslides during hurricane Faye, which struck in September 2008. Despite the lack of erosion mitigation, farmers continue to plow the sloping hillsides revealing red rocky soils that farmers report are continuing to decline in fertility. The crop mixture grown here, millet and sugarcane grown alone, and maize, pigeon peas, peanuts, sweetpotatoes, and manioc grown in some combination simultaneously, was similar to that of the other rain fed areas. However, the severity of the erosion was more acute than in the flatlands, so yields are likely not as high in this region. No irrigation w as available in this area.
90 Lakou. In rural Haiti, the kitchen garden, located in the yard ( lakou) surrounding the living area11, is one of the most important sources of nutrition for the household (Correia 1998). The lakou is a unique agricultural habitat that is present within each of the regions mentioned previously. The majority of the lakou observed were managed agricultural landscapes that maintain high species diversity and provide food, fodder, fiber firewood, and botanical pharmaceuticals to the household. In a typical lakou, the p erimeter of the house, outbuildings and walkways consist of bare dirt which is swept clean on a daily basis. Thus any organic matter, leaves, and dust accum ulate in the soil of the cultivated portion of the lakou. Several animals, including free range chickens, guinea fowl, tu rkeys, pigeons, pigs, goats, and dogs sleep and may be fed in the lakou. Their waste along with kitchen scraps, dishwater, and agricultural byproducts provides nutrients, moisture and organic matter to the soil. The lakou is bordered by living fences, pr imarily rakt mixed with various tree species and gated which offers protection for animals and highly valuable crops Ornamental plants and borders were a common feature of urban lakou and less common where homes were more dispersed. Ma ny farmers wer e exploiting the increased fertility of the lakou by planting crops like millet, cassava, and pigeon pea near the house. Others planted various fruit trees (Table 41) i ncluding those that were grafted. Many of the crops (Table 43) within the lakou, inclu ding beans tropical pumpkin, passion fruit vines, and castor bean (which is used medicinally and as a hair treatment), as well as various edible weed species appear to be volunteers that are selectively maintained. Papaya, plantains and bananas, and root crops including cocoyam and taro are planted in various 11 A typical lakou usually has 12 bui ldings for living quarters, a kitchen, outhou se, and sometimes a granary.
91 microclimates within the lakou. Since seeds are not readily available in the market or are cost prohibitive, most of the vegetables and edible leaves in the garden are volunteers from kitchen was te or propogated from tuber propagules or cuttings like sweet potato, chaya ( Cnidaria esculenta) and Haitian Basket Vine ( Trichostigma octandrum ) Chaya, Haitian Basket V ine, and Moringa oleifera were observed in the lakou in the fl atlands and in Skadi, where development projects have promoted these highly nutritious and drought tolerant plants. These species provide an additional source of calories and nutrients year round including during the leaner seasons when dietary staples ar e not ready to be harvested. Vegetables like cabbage, onions, carrots, and tomatoes were not observed in the lakou vis i ted. The plants found in the lakou reflect the preferences of the household, labor availability, and household income. The degree to w hich the lakou was mai ntained and the number of cultivated plants in the yard appeared to be related to the households available labor and income. Weeds were a salient feature of the lakou and varying levels of weed management were practiced. Because many weed species a re used as animal forage and some are consumed by the household, weeds were never completely eliminated from the lakou. In households that included two ablebodied adults or older children, the lakou was usually well kept and provided significant s ource of food for the family. Single parent households with young children were typically among the most economically disadvantaged and had fewer edible plants and landscaping features such as ornamental plants in the yard. Thus, the lakou might serve as an economic indicator for this region.
92 Training The majority of the individuals interviewed fit Wilken s (1972) description of cultivators whose knowledge and methods are derived from individual and social experience and who use only locally av ailable energy and materials In rainfed areas, farmers were not using external agricultural inputs, but that was not always the case. During the 1970s 1980s fertilizer was more readily available in this region, and some of the individuals interviewed m entioned that they used to receive fertilizer and training from a local mission group called AEM12 that previously operated a well funded agricultural program. Most of the interviewees stated that their fathers taught them to farm, although a few mentioned training from an agricultural extension agent ( agrinom ) at some point during the past 30 years. The choice to employ the methods and knowledge handed down may be a consequence of the lack of agricultural inputs in the area and insufficient cash to purchase what limited inputs do exist. One farmer reflected that he learned to plant field crops in rows from an agrinom during the mid 1970s, yet he does not follow that recommendation anymore. An agri nom we interviewed stated that farmers have been taught to plant with teknik agrikl (planting in rows) to increase yields, but they are not motivated to use this technology. He is of the opinion that lack of dedication and time are the main reasons that farmers in this region are not successful. 12 AEM is the French abbreviation for Association Eglise Missionaire Missionary Church Association (MCA), which founded a mission in LaPila during the early 1950s. The property is now run by the Haitian national board of AEM based in Port au Prince, and a local committee. An active dispensary managed by a local Haitian doctor and a church operate from the property. The AEM national conference, various camps, and retreats are held there as well. Meeting rooms, houses, dorm s, and a dining hall are available for groups to rent for retreats, conferences, and meetings.
93 A handful of interviewees had received training through employment with NGOs including PADF (PanAmerican Development Foundation) and World Vision. Others had participated in formal training programs ( formasyon) through the Haitian American Friendship Foundation (HAFF), World Vision, and Mouvman Peyizan Papay (MPP) an affiliate of La Va Campesina13. One farmer who had worked at a World Vision nursery had been trained in compost making and basic pest management. Prior to 1988, a Canadian agricultural mi ssionary Agrinom Richard, would plow peoples fields with a tractor. Some farmers mentioned that getting a tractor back in the area would be helpful to them. Challenges to Agricultural Production Farmers were not hesitant to share with Sondeo teams the struggles they encounter growing food. Yield declines, waning soil fertility, seed and input inaccessibility ( especially fertilizer ) pest problems, unreliable rains, inability to pay hired laborers, and the absence of irrigation infrastructure were pres ented to interviewers as some of the most significant factors limiting food production in the study area. Farmers identified February May as the busiest time of the year and February, April, and J uly as some of th e months when their families are most likely to be hungry. Many people eat the seed that they had saved for planting and have neither seed nor the resources to purchase new seed when it is time to plant. Pest management 13 La Va Campesina, established in 1993, is a multicultu ral movement of peasants, small and medium sized producers, landless people, indigenous people, rural women and youth, and agricultural workers whose principa l goals are to develop solidarity and food security ( http://www.viacampesina.org ).
94 F armers in this region have had little to no formal training in pest management, but practice traditional methods and experiment with locally available materials (Table 44). Other pest management information is most often obtained by the agrosupply store where they purchase seeds and chemicals or from the individuals selling chemical pesticides in the market. Many farmers were familiar with common insect pest species (Table 45) yet none of the individuals we spoke to were aware of beneficial insects. When asked whether there were any good insects at work in their garden, most farmers responded with laughter. Only one farmer mentioned that he had observed an insect attacking a caterpillar, but did not know what it was. Others reported that birds including krabye dlo (an egret species) and snakes eat insects and snails in the garden. These same birds, along with the weaver bird, become major pests prior to the November and December millet harvest. In the flatland, hillside, and s avann regions, we found that, in general, farmers were too poor to use chemical pesticides on their crops. None of the farmers of agronomic crops shared pest prevention techniques with us. When asked how they control pests, one farmer said he does not have the money for pesticides, so he picks all of the caterpillars off his vegetables by hand. Another farmer lamented that, He plants maize. Pests eat his plants, and so he plants again until he has no more seeds. The overarching sentiment was that because they cannot afford pesticides, they cannot control or prevent pests from stealing their crop. Although the farmers we interviewed took no overt precautions to prevent pest problems, the biodiversity maintained both in the lakou and in their gardens through extensive intercropping, crop rotation, and living fences, and the lack of broadspectrum chemical use likely reduces some pest pressure (Altieri 1994).
95 Vegetable growers in the irrigated lowlands of Lt bBohoc were using pesticides on their cabbage, which is their highest value c rop. Because cabbage is intercropped with a variety of other vegetables, like tomatoes, okra and peanuts, the other vegetables may get sprayed as well if there are visible pests or damage. Pesticide application regimes varied among interviewees. One far mer who was employed by a NG O tree nursery received pesticide training through his work. He applies Malathion (an organophosphate), or Sevin (a carbamate), to his cabbage crop at two week intervals but stressed that he only applied the product when he sees pests. This indicates that he performed some level of monitoring. Other farmers in Lt b Bohoc were using a diazinon product called Kayazol, which is purchased by the bottle as a concentrate and diluted with water. This pesticide is applied on a t hreeday spray calendar per the instruction of an agrosupply vendor who lives in Hinche and is their source of pest management information. They have been following this regime for the past 2 years and indicated that they have not seen resistance. Howev er, when the pesticide fails to kill all the insects they use a more concentrated application. Kayazol was selected by these farmers because it is the least expensive and effectively prevents losses from aphids, cabbage root maggot ( Delia radicum ), and c abbage moth ( Pieris rapae ) when cabbage is planted in early fall (late August September). Farmers plant early to avoid infestation by the diamondback month ( Plutella xylostella ) which becomes the key pest of cabbage beginning in December through the wint er growing season. Diazinon does not provide adequate control of diamondback month in this region, so a product called Match is used following the same 3day spray calendar. Because the label was unavailable, and there are two products known for use in cabbage with a similar name, it is unclear which
96 chemical these farmers are using14. Match is twice as expensive as the diazinon product, so planting early has tangible economic benefits. Some farmers in the vegetable growing region have had training in the production of biorational pesticides at MPP including neem w ater ( dlo nim ) made f rom the leaves of the neem tree, Azadirachta indica hot pepper water ( d lo piman ) Capsicum annum and Sour orange water ( dl o zoranj sik ), Citrus reticulata In general, the biorationals were deem ed effective against young caterpillars however, the control of late instar larvae requires synthetic chemicals. The labor associated with making biorational pesticides was stated to be a deterrent to their use. Additionally, d lo piman can damage the crop if the formula is too concentrated. F armers did indicate that when they use a natural product, it frees up money for seed and fertilizer. However, farmers associate reduced risk and less labor with the synthet ic chem ical pesticides purchas ed from the a grosupply store. H erbicides are not employed in this region yet, although one farmer asked whether the team knew of a pesticide that would kill nutsedge. He had heard about herbicides from the agrosupply store, but had not yet purchased them. Hand weeding, plowing, and crop rotation are the most common weed management practices. Chemicals are applied using pump sprayers when they are in good repair. Spraying is the preferred method for chemical application; however, pump sprayers are difficult to find and maintain. One farmer applies chemicals using a coffee canister with holes punched in the bottom and others use a watering can, although they stated that using a sprayer i s a more effective way to deliver the produc t and kill pests. Extension 14 Match EC 50 is a product used throughout the Caribbean and South America. Mattch TM is a reduce risk pesticide derived from the bacteria Bacillus thuringiensis (BT). Neither is recommended for use every 3 day s
97 personnel and vegetable growers, including those whose primary source of pest management advice is the agrosupply store, described the protective clothing that should be worn during pesticide application: long sleeves, pants, gloves, a hat, a handkerchief over the mouth, socks and shoes or boots, sun glasses The majority of these pesticides used on vegetable crops in this region (Table 4 6) are broad spectrum chemicals with known nontarget effects on beneficial insects, bees and fish. However, many of them have relatively low mammalian toxicity, short residual activity, and a 12hour reentry interval A pplication areas are not labeled and children run barefoot through the gardens. Longterm risks associated with pesticide exposure and increased dangers for children and pregnant women are likely not being emphasized by vendors. None of the interviewees mentioned the potential harmful effects that chemical pesticides can cause to natural ecosystems, including water resourc es, wildlife, and domesticated animals. It is important to note that the irrigated lowland is within the flood plain of the Bohoc R iver. Runoff from these plots flow s directly into the river during the heavy rains associated with the wet season and hurri canes. The proximity of the river, open irrigation system, and suite of chemical pesticides known to have deleterious effects on aquatic life, natural enemies, and birds should be a concern of the NG O s promoting agricultural and environmental education, especially those providing resources for purchasing chemicals. Diazinon in particular has restricted use registration in the US because of its negative impact on birds (Eisler 1986; Extoxnet 1996).
98 Seed conservation Although few farmers in this region are currently using pesticides on their crops, when asked how they conserve their grain, most interviewees indicated that they store seeds using synthetic pesticides. Many informants referred to this chemical as DDT15, while others described it as a cream colored powder they purchase by the spoonful in the local Bohoc and Pignon markets. This was later confirmed by a visit to a pesticide vendor at the Bohoc market. The powder sold was stored in a Tupperware container; however, the product package was available for the customers to view. The labeling on the package was fairly worn and written in English and Spanish, so it is highly likely that neither the salesperson nor the customers were able to read any of the instructions. It was unclear whether the powder sold was diluted with flour or another filler agent, a practice that occurs in other regions of Haiti. We observed a pesticide sales transaction at the Bohoc market where the vendor rolled a piece of a paper cement bag into a cone, scooped 4 small teaspoons of the powder into the homemade packaging, which he folded up and sold to a woman for 1 Haitian dollar (roughly $0.13 USD). Directions were not offered to the customer, nor was the package labeled or properly sealed. When the investigators ask ed how to use the product, he said to dry the seed in the sun, mix in the powder, and store the grain in a drum. He said that the powder is not dangerous, and that seeds preserved with the pesticide can be eaten without even washing them. A passerby interjected that the seeds must be washed before they are consumed. Others disagreed, saying that the 15 The two agricultural extension agents interviewed explained that many peasant farmers in the region use the term D DT and pesticide synonymously, but that DDT had not been used in the region for several years. During the Duvalier regime, houses in this region were sprayed with DDT in an effort to abate mosquito borne illnesses.
99 pesticide should only be used to conserve seeds intended for planting in the next season, not for grain that will be eaten because pesticides are not good for the body16. During interviews away from the market, the same method for acquiring and using the pesticide was described. Opinions regarding the safety of consuming treated seeds varied. One informant made an apt summation, you shouldnt eat seed saved with pesticide, but when youre hungry and have nothing else to eat, what else can you do? Haitis hot, humid climate is not ideal for seed storage. Stored product pests ( timix ) and mold ravage grain stores, and temperature and humidity fluctuation decrease seed viability. Several farmers mentioned that they would plant more if they had more seed. Others use ineffective storage methods that result in loss of seed quality and volume to pest and disease infestation. According to the agrinom interviewe d, seed saving is one of the most limiting factors to production in this area. He explained that people sell everything at the harvest to get cash when the prices are low and then either do not have seed for spring planting or are forced to purchase their own seed back again at higher prices. Farmers shared that vegetable seeds were difficult to access and very expensive. The agricultural supply stores that provide vegetable seeds ar e in Hinche and to a lesser degree in Pignon, and seeds selection and av ailablility is fairly limited. Some of the informants shared their interest in developing methods for saving and storing vegetable seed. Farmers employ multiple methods to store grain for future consumption or planting with varying degrees of success. The traditional method for grain storage involves weaving the outer maize husks into a long vine ( lya nn mayi literally vine of maize), 16 This is a direct translation from the kreyl, pa bon pou k.
100 which is hung from a tree post laid on the roof of a house. Cobs are stored on the vine and removed as needed for household consumption. Grain that is removed from the cobs is stored in sacks or drums, often mixed with a natural or synthetic product, to be used for planting in the following season. Farmers also conserve seeds using tablets of pesticide available at the agricultural supply store in Maissaide (a 10hour walk from the study area, requiring crossing the Bouyaha River ) or Hinche. Encapsulated pesticides including aluminum phosphide and Lindagrain (Lindane), are wrapped in cloth and added to a sack or dru m containing grain and stored in a kanari, a raised silo that looks like a tree house, or inside the house. Additionally farmers utilize natural products for seed conservation including ash and crushed limestone ( tash) with varying degrees of success. I n general, seeds that have been conserved with a powdered chemical like DDT are not consumed by the household unless they are desperate, while the seed preserved using natural products or wrapped tablets are considered safe for consumption. H owever, opi nions and experience with pesticides for seed conservation varied among the interviewees. Disease m anagement Cultural control methods are often so closely associated with management of an agricultural landscape that they are sometimes overlooked as pest and disease control strategies (Flint and Gouveia 2001). Crop rotation, which aids in the prevention of soilborne diseases, pests, and nematode problems, as well as intercropping are routinely practiced by many farmers throughout the study area. In the hot humid tropics where fungal diseases thrive, a polyculture system that is not spatially arranged in densely packed rows likely increases aeration and reduces the conditions that create plant disease outbreaks. Multiple crops in a given area may help prev ent disease
101 epidemics as well. However, l ack of fertility and drought stress likely contribute to weakened plants that are more susceptible to secondary pathogens and plant disease. No fungicides are currently employed for disease management in the study area. When asked about disease problems, farmers in the flatlands, with the most extensive sugarcane farming, menti oned chabon17, an infection of newly emerging shoots that turns them black. Chabon is managed by cutting back the sugarcane or burning. It was not clear if there was a certain time of the year that this disease was most prevalent. In the same area, interviewees described a white chalky disease on the cane stalk. They did not attribut e severe losses to this disease, although infected cane is weaker, and they tolerate its presence by scraping the cane with a machete. Upon observation it appeared that the cane was infested by a white scale insect rather than a plant disease. Farmers als o described a black fungus on the seed heads of millet that is a problem during rainy winters and makes animals sick. A typical lakou contains papaya plants of various ages, and most often there is at least one severely stunted papaya plant displaying Pa paya Ringspot Virus ( P R S V ) symptomology: malformed chlorotic leaves with mosaic patterning, tall spindly habit, shrunken fruit that senesces prior to ripening, and ringspot markings on the fruit. PRSV is an ap hid vectored plant virus that limits producti on throughout Florida and the Caribbean basin (Gonsalves 1994). When asked about pest and disease problems, several informants pointed at their papaya plants and described various insect pests that they had seen on the papaya, but none of them mentioned v irus being the cause of the disease. Consequently, several lakou have an infected tree that serves as a 17 In Kreyl this term is also used to describe charcoal and anthrax disease.
102 reservoir of PR S V readily transmitted by aphids to healthy plants in their yards and their neighbor s Soil amelioration and plant nutrition S oil f ertility was one of the greatest concerns of all of the farmers interviewed. Several farmers lamented that yields have been declining in recent years and associate losses with poor soil. One farmer specifically mentioned that the number of kernels on an ear of maize is much less than it used to be. He attributes this to poor soil. Incomplete kernel set can be caused by drought and lack of nitrogen (Nielsen 2009). An octogenarian said that when she was first married, one marmit of maize seeds would yield four lyann mayi However, now they are getting one lyann mayi from the same amount of seed. Now she thanks God i f she gets back what is planted. None of the farmers encountered in the flatland, hillsides, or savann were currently using chemical fer tilizer or applying other soil amendments to their field crops. Fallow and crop rotation were the only methods used to improve soil fertility in rainfed growing areas. In Terre Glise, farmers had previously purchased or received fertilizer from an AEM agrinom but it is no longer being provided and is too expensive or simply unavailable in the market place. There is an understanding that additional fertilizer would improve production, yet none of the dryland farmers interviewed mentioned using any other nonchemical soil amendment on their field crops. One innovative farmer who has used fertilizer in the past, said that if he still had fertilizer, the food produced in his lakou would be sufficient, and he would not have to farm any of his other land. Another man who lived in the Savann and worked for a World Visions nursery made compost from sugarcane bagasse, which he used in his lakou.
103 Vegetable growers and those who had received training through employment with a NGO were taking the most initiative to conserve and improve soil fertility among all of the farmers interviewed. In Lt b Bohoc, we encountered individuals who were using bat guano that they purchase by the donkey load from caveowners in the mountains above Skadi. These farmers also grow jicama as a green manure that they incorporate into the soil along with other plant debris as they plow their gardens. Many of these farmers burn charcoal in these gardens because the residues bring additional nutrients to the soil, and they believe the ash c an serve as a deterrent to snails. The vegetable growers purchase urea for use on young cabbage and another fertilizer called ko mplimant which is applied during cabbage head formation. After the first cabbage harvest, stubs are left in the ground to r esprout, making smaller heads. However, no inputs are appropriated to this bonus harvest. Vegetable growers described removing maize stalks and cabbage roots following the harvest to prevent them from stealing nutrients from the soil. Although many interviewees were were familiar with compost making, none were producing their own. Household Values and Goals In this region of Haiti, a household unit is best defined using the Kreyl term lakou, and includes all of the people who live in the house or in bui ldings that share the same yard. This may encompass parents, children, grandparents, aunts, uncles, cousins, nieces, nephews, and children that serve as household domestic s ( restaveks18); all of whom share a single cooking fire and work together to achieve household goals. The 18 The word, restavek derives from the French words, rester to stay, and avec with Restavek children stay with a family in exchange for assistance with household duties Many consider this a form of modern slavery.
104 size and composition of a lakou varied greatly among interviewees, and i t was not uncommon to find large families in the north central plateau of Haiti. H ousehold composition determines the amount and kind of labor available to work toward the households production goals as well as the amount of food required to meet nutritional needs M ore mouths to feed often means less food, or less nutritious food, on each plate. I f the re are several children in the lakou, the household will incur additional costs for tuition, uniforms, school supplies, and books Neither household composition nor goals remains static. For example, as children grow, they begin to participate in reproduc tion activities or require less supervision, thereby liberating the female head of household to engage in more production activities. Adolescents, especially males, function as adults and may boost production activities if they are not enrolled in school. Food consumption and nutrient requirements for families change over time as well (Hildebrand and Sullivan 2001). For subsistence farmers, consumption and production intersect at the level of the household, and the means of achieving a livelihood is ine xorably connected to the household unit (Litow 2000). Production activities of household members on and off the farm are not simply business ventures targeting maximum profit and returns. The farm represents a household first and is a means for achieving a set of goals (Hildebrand et al 2001 ). Household goals are met when levels of resources (land, labor, equipment), food consumption, and cash for required expenditures ensure food security and then maximize the amount of cash available for discretionar y spending (Hildebrand et al 2001 ). In rural Haiti the households primary goal is to gather enough cash to send children to school. One farmer in Terre Glise said, We sell some of our maize even
105 when were hungry to pay for school. Education is highly valued in the communities visited during the Sondeo. Multiple families suggested that if the children are enrolled in school they are not involved with hard labor. Rather, it is their job to study, so that they can have a better life. Having children enrolled in school appears to be a good economic indicator for this region. Recommendations In general, we propose that rural development programs consider the demands of the agricultural calendar, especially during the months when farmers will be the most or the least occupied with demands that contribute to household production. Households are busy preparing the land and planting March through May. This would be a poor time to host an intensiv e educational program Additionally, the households fiscal and hunger calendars should be considered since there are months when families are cash poor and other times when they anticipate making a large purchase or paying back a loan. Farmers are m ost desperate for cash during July when they are forced to eat from the market because few crops are ready for harvest. School fees are due in September and January which are significant times of financial stress Households anticipate cash shortages when they can, by raising animals and selling when the bill is due. However, illness, or inclement weather often thwart the efforts of rural farmers to make ends meet. Regarding any proposed intervention, we encourage the reader to reflect on these two sta tements: If its such a good idea, why arent they already doing it?" and Will the household reap short term economic benefits? In general, farmers are risk averse (Lundhal 1984) and hesitant to adopt new technology that is labor intensive (especially during the busy season) without foreseeable benefits in terms of increased yields in the current growing season, decreased production costs, additional food, or increased
106 income (Shannon et al. 2001). Labor is a major constraint to production for peasant farmers (FAO 2001). A lthough unemployment is rampant in Haiti, the lack of jobs cannot be misconstrued to mean that an individual household has unlimited labor available to them to meet their household goals Seasonal l abor and the chronic shortage of cash are major constraints to subsistence farmers in rural Haiti so management techniques that are cheap, labor reducing, and effective are priorities for farmers Those that offer short term economic benefits are more likely to be accepted. During a USAID funded productive land use systems project in Haiti, Shannon et al. (2001) found that farmers were more likely to invest in the establishment and maintenance of erosion control hedgerows if they provided short term economic returns or food f or household consumption and if they had participated in an educational program on soil conservation. Through their conversations with local farmers a new concept for erosion mitigation emerged called b ann manje which is a play on words in Haitian Creole, meaning both strip of food and a lot of food (Freeman and Laguerre 2002, Shannon et al. 2001). By encouraging farmers to establish a hedgerow of highvalue perennial or longlasting annual crops -literally a strip of food-and planting along the contour of hillsides, growers increase land use efficiency and mitigate erosion while contributing to household production (Shannon et al. 2001). Community Associations Working though existing community associations ( gwoupman) can be an effective method for promoting sustainable agriculture practices through training programs, information and idea exchange, and planting material distribution. Established gwoupman also provide a forum to gauge community support and interest in various programs. Unfortunately many of the poorest individuals are not able to participate in a
107 gwoupman because they do not have the money for dues. Without a dues paying membership, several gwoupman fold because members perceive that without money for projects, they can do very lit tle. Making grants or microloans available for gwoupman projects may reduce the need for dues to be paid and have the benefit of increased gwoupman membership, especially among the poorest individuals in the community. We also observed that many gwoupman collapsed when their leaders passed away. Setting up a gwoupman leadership team or model that continually trains leaders and mentors future leaders may keep gwoupman from dissolving and may improve continuity of association activities. Focus on the Lakou Declining soil fertility, inconsistent rainfall, and lack of irrigation are some of the key factors that limit agricultural production in this region. The lakou generates organic matter and has the greatest potential for rapid soil improvement, compost making, and increasing production in the short term. Labor by multiple household members is readily available because no travel is involved. There is great potential for developing projects for the lakou focused on increasing agricultural production incl uding specialty crops, and underexploited food crops that extend the harvest season. These projects would be applicable to all age groups and especially to young people. The lakou, which is primarily maintained by women, is believed to be more important for ensuring household food security than field crops (Correia 1998) and could potentially serve as an economic indicator for this region. During two interviews with men, the investigators asked about chaya plants growing in the yard and it was clear that the men had no idea they were edible plants. At one of the houses, the wife of the
108 man being interviewed explained how she used chaya and that she had received cuttings from participating in a training seminar at a local NG O Extend the H arvest Househ olds identified February, April, and July as the leanest months of the year. Although some interviewees were reluctant to admit that their family was ever hungry, some of the more candid informants explained how households get by when crops are still in t he ground but not ready to be harvested. The agricultural extension agent we spoke with recommended people plant cowpea at the beginning of the rainy season to avoid summer hunger. Cowpea is highly adaptable tolerating both marginal soils and dr ought ( Eh lers and Hall 1997) Additionally, it is a legume t hat is high in protein, provides nutritious leafy forage for animals (Kitch et al. 1998), fixes nitrogen, and may suppress weeds, nematodes and other pathogens ( Wang et al. 2006). Early maturing varieties can be harvested 6070 days after planting (Ehlers and Hall 1997). NGO s and agricultural researchers should pursue the potential of cowpea in this region, with local and improved varieties Superior and underexploited fruit tree cultivars may further extend the harvest during lean summer months as well. Erosion M itigation The barren hillsides and patches of bedrock cropping up in gardens are reminders that soil loss through erosion is the greatest threat to the loss of productiv e land in the area. Nevertheless, most farmers in this region use animal traction to plow, even on hillsides. Erosion control, through notill agriculture or reduced tillage, could save some of these farmers money in plowing and labor costs and would reduce the rate of soil loss. No till systems increase soil moisture, improve soil structure, and can increase soil fertility over time as well (Powers and McSorley 2000). There are constraints to a
109 notill system especially where no herbicide is available. Despite the effort of several agencies throughout Haiti to promote the construction of erosion control berms, few farmers in this area employ erosion mitigation on their own land at their personal expense. The erosion control barriers observed were funded by an international development organization. This region might benefit from experiments with bann m anje. Integrated Pest Management T raining Many of the pesticides employed by vegetable growers in Lt B Bohoc are broadspectrum pesticides (Table 4 6) that indiscriminately kill pests, natural enemies, and pollinators and have detrimental effects on fish and birds. M any of the products described by farmers have relatively low mammalian toxicity and low residual activity However, highly toxic pesticid es have been used in the southern vegetable producing region of Haiti (Bishop 1995). In 1987 and 1988, the neighboring Dominican Republic had the highest rate of illegal pesticide residues on foods imported into the US (Murray 1994). At any time, more toxic chemical products could enter the local markets from border towns or larger cities. Therefore, equipping farmers with sustainable alternatives and knowledge should be a development imperative be fore the potential for exposurerelated illness and other nontarget effects on human health and the environment increase. Broadspectrum pesticides kill the beneficial organisms that perform key ecosystem functions including the regulation of pests, decomposition, and pollination (Lewis et al. 1997, Altieri 2002). Without the stabilizing presence of natural enemies, pest populations can surge to higher levels than prior to the chemical application (Lewis et al. 1997, Flint and Gouveia 2001). The ecological effects of chemical based pest
110 management strategies create a built i n need for more pesticide use ( Murray 1994). This phenomenon is described as the pesticide treadmill ( Van d en Bosch 197 8); once a farmer starts using broad spectrum agricultural chemicals, they eventually have to increase chemical rates and/or applications to achieve results similar to the first application. Applying the same chemical on a 3day spray calendar is a proven formula for developing resistance in at least one of the areas key pests, Plutella xylostella Farmers using the spray calendar have already initiated their walk on the pesticide treadmill and described increasing pesticide rates when lower doses fail. In other regions of Haiti, chemicals like Malathion and Sevin are no longer effective for key pests of cabbage and maize. Reliance on agricultural chemicals as the primary method of pest control is not a practical longterm pest management strategy, especially for low resource farmers in developing countries like Haiti, where poverty, illiteracy, and limited access to training compound the risks associated with pesticide use. Moreover, dependence on agricultural salespeople as the predominant source of pest management information is a conflict of interest that is likely to be biased toward the end goals of the agrosupply store. Peasant farmers in developing countries need sustainable alternatives and the opportunity to learn about pest management techniques that do not rely on pesticides. Integrated pest m anagement (IPM) is an ecologically based approach to pest control that combines the commonsense practices of local farmers with scientific knowledge to reduce the risk of pest outbreaks while minimizing the negative impacts of pesticides on the environment ( Luckmann and Metcalf 1994, Geier 1966). IPM encourages a multi pronged preventive approach to pest control by simultaneously employing multiple crop husbandry practices that disrupt pest life cycles or conserve
111 and enhance the regulatory activity of natural enemies ( Luckmann and Metcalf 1994, Geier 1966) In an IPM system, p esticides are not the first line of defense, nor are they applied on a schedule. Rather, farmers are encouraged to regularly monitor their crop and employ a chemical control method as a last resort. When pesticides are used, least toxic methods that have few nontarget effects are recommended. Products like oils and soaps, which are nontoxic, effective in managing certain pests, inexpensive, locally available and lack lengthy preparation time and processing, could be easily integrated into a pesticide rotation program in the vegetable growing region. Although some sustainable alternatives are available, synthetic pesticides are the preferred choice of local farmers. IPM training would provide a sustainable alternative to the pesticide calendar and has proven potential for reducing costs and risks to field workers, their families, and the environment. In their work with farmers in Honduras, Wyckhuys and ONeil (2007) report ed t hat IPM training strengthened farmers appreciation of natural enemies. Farmers who understood the role of natural enemies within the agroecosystem were less likely to use pesticides and more inclined to employ more chemically benign control measures (Wyckhuys and ONeil 2007). Additionally, increased understanding and appreciation for the work of natural enemies within the agricultural landscape can lead to farmer innovations in pest management. In an IPM training program for women in Latin America, tra iners provided information about IPM principles and pest and beneficial insect lifecycles, rather than teaching general pest management (Meir 1999). Participants were encouraged to experiment on their own farms and develop locally appropriate technologies (Meir 1999, Bentley 2003). During this education program, one student began spraying sugar water to attract wasps and ants to her gardens to eat
112 caterpillars and other pests; this innovation has proved successful and has been promoted by the women in the program (Meir 1999, Bentley 2003). Currently farmers in the greater Bohoc region lack the knowledge to recognize and understand the role of beneficial insects within the agroecosystem. T he farmers we interviewed in the irrigated lowlands were eager for information, and requested to join in any upcoming training programs offered on vegetable production. Partnershi p with local and international NG O s and wildlife conservation groups to develop and fund IPM training specifically regarding the role of natural enemies could fill an important educational need in this region. Target Future Farmers Children are particularly vulnerable to the consequences of environmental and health hazards linked to pesticide misuse ( Goldmann 2004 ). Nearly half of Haitis population is under 18 years of age (CIA 2009) Young farmers appear to be more conservationminded than older farmers and more inclined to adopt sust ainable practices especially after participating in training (Shannon et al 2001) Equipping the future farmers of Haiti with knowledge and experience with IPM technology will likely improve local food security, prevent further degradation of Haitis del icate environment, and prepare a generation of opinion leaders committed to the conservation of natural resources through sustainable agriculture. Conclusions The Sondeo proved to be an effective and appropriate method for obtaining information for this introductory study. Visiting people at their homes for conversation is an important part of Haitian community life. W e found that many farmers were eager to share with the teams their gardens and their struggles to help us understand their
113 culture and values. Our goal was for farmers to serve as informants about their experiences and opinions through informal dialogue. Although every discussion included the same key topics, the questions differed in both the order asked and the phrasing depending on the context of the exchange. The result is that the data collected are not quantifiable statistically. However, the farmers shared information that may have never been garnered through a traditional survey with predefined questions. The Sondeo format is a dynamic process, and as the teams grew more familiar with directing conversations to meet the goals of the study, the questions changed and new discussion points arose during our nightly meetings. For example, one of the goals of the survey was to underst and farmers experiences using pesticides. During the first day of interviews every team asked farmers how they controlled pests in their gardens. All interviewees said they did not use pesticides because they are too expensive. However, when one group s conversation about the challenges to growing food led to the topic of seed conservation, the farmer shared how he used a chemical pesticide to conserve his seeds. When the team met that evening to discuss the information gathered throughout the day, we learned that to get pesticide information we would need to ask about seed conservation during subsequent interviews. The Sondeo illuminated the plight of residents within the study area as well as the unique set of strategies they employ to provide for their fami lies. Agriculture is the princi p a l economic activity in this region, but farms are generally small, inputs and seeds are difficult to access, and soil fertility is declining rapidly. Harvests are not abundant and households are not producing enough to meet their consumption needs while conserving seed for the next growing season. Dietary protein is primarily limited to legumes because animals are primarily kept as a form of savings account. There is a
114 dearth of employment opportunities in the Cen tral Plateau and residents face a chronic shortage of cash. Despite these serious limitations, rural Haitians persist with remarkable ingenuity. Education and training, specifically in IPM and soil and nutrient management may help Haitian farmers obtain more success and food security while conserving existing natural resources.
115 Table 4 1. Fruit and nut trees found in greater Bohoc region. English Kreyl Latin name Cashew nwa kajou Anacardium occidentale Pineapple anana Ananas comosus Soursop kowos l Annona muricata Custard apple kashimankb f Annona reticulata Sugar apple kachima n Annona squamosa Bread fruit veritab Artocarpus altilis Breadnut bwa pen Artocarpus spp. Starfruit kar ambola / fwi zetwal Averrhoa carambola Papaya papay Carica papaya Grapefruit chad k Citrus .x paradis Tangerine 1 mandarin Citrus tangerina C. tangelo Key lime sitw on Citrus aurantifolia Lemon l imon frans Citrus limon Sour orange zoranj si Citrus spp. Tart orange s ir t Citrus spp. Sweet orange zoraj dous Citrus sinensis Coconut kokoye Cocos nucifera Coffee k afe Coffea spp. Mango 2 m ango Magnifera indica Spanish lime k en p Melicoccus bijugatus Plantain/banana banna n n/fig Musa spp. Passion fruit g renadj a Passiflora edulis Avocado z aboka Persea americana Guava g wayav Pisidium guajava 1 Includes tangeringes, tangelos, & mandarins 2 Haiti boasts hundreds of mango varieties, some of the common varieties in this region include: Batis, Dou Dou s Fileb lan, Fransik, Kodok, Ti Rouj, Gp, and Savon which is the latest bearing variety in the region.
116 Table 42. Field crops cultivated in rainfed areas of the greater Bohoc region. English Kreyl Latin name Cereal Grains Corn m ayi Zea mays Millet gwo pitimi pitimi pitimi novanm ? 1 Sesame j ijiri Sesamum indicum Rice (upland) d iri Oryza sativa Roots/Tubers Cassava, bitter many k si Manihot esculenta Cassava, sweet many k dous Manihot esculenta Cocoyam mazonb l Xanthosoma sp. Sweetpotato p atat Ipomoea batatas Taro tayo Colocasia esculenta Yam yam sigin, yam gine yam reyal Dioscorea spp. Beans/Legumes Peanuts pistach Arachis hypogaea Pigeon pea pwa kongo Cajanus cajuns Local bean cultivars gwo pwa 2 Phaseolus spp. Lima bean pwa boucousou/ pwa chous Phaseolus lunatus Black bean pwa nwa/ pwa rache Phaseolus vulgaris Cowpea pwa kouri Vigna spp. Black eye cowpea pwa je nwa Vigna spp. Cover crop legumes Jicama pwa many k Pachyrhizus spp Velvet bean vevet bin Mucuna pruriens Jack bean jak bin Canavalia spp. *Indicates that this crop was also observed in irrigated fields 1 Kreyl translators used the English word m illet to describe a crop that appeared to be sorghum ( Sorghum spp.). Three local varieties are grown in this region. 2 The term gwo pwa includes several local bean varieties including but not limited to pwa chikan ame, pwa kaka chat, as well as limabeans and black beans.
117 Table 43 Vegetables reported by informants to be cultivated in the study area. English Kreyl Latin name Vegetables l egim Okra gon bo/kalalou Abelmoschus esculentus Scallions p wawo Allium spp. Onions z onyon Allium cepa Spinach zepina Amaranthus spp. Beets b t rav Beta vulgaris Cabbage, head chou Brassica oleracea var. capitata Cabbage, Chinese chou fey Brassica rapa Pepper, hot piman pik e Capsicum annum Pepper, sweet piman dous Capsicum annum Watermelon melon dlo Citrullus lanatus Chaya chaya Cnidoscolus chayamansa Pumpkin joumou Cucurbita moschata Carrots kaw t Daucus carota Tomato t omat Lycopersicon esculentum Moringa, leaves d oliv Moringa oleifera; M. stenopetala Chayote squash m iliton Sechium edule Eggplant berej n Solanum melongena Haitian basket vine lyann panye Trichostigma octandrum
118 Table 44. Pest management techniques as reported by informants. English Kreyl Crop Man agement s trategy Ants Foumi Corn Millet Peanuts Vegetables Coat seeds in castor bean oil before planting Plant with ground castor bean, Ricinus communis Plant with the leaves of ogoun ? Synthetic pesticides 1 Aphids Picho n Citrus One farmer tried ash mixed with water, but found it in effective. Caterpillars Cheni Corn Millet Cassava Vegetables Hand pick caterpillars Synthetic chemical pesticides ( Table 4-6 ) Natural pesticides: dlo piman (h ot pepper water) & dlo nim (n eem water). Planting cabbage early to avoid heavy caterpillar pressure later in the season. Weevil Tiyogann 2 Banana Plantain Planting in pine sawdust (pines are no longer found in the area) Moving planting material to a new site Planting in ash Snails Kawasol Vegetables Mechanically remove them from diurnal resting habitat Ash from making charcoal Post harvest p ests Timix Beans Corn M illet Selecting corn varieties with tight ears Synthetic pesticide: cream colored powder; DDT; Lindagrenn Ground l imestone 1 According to the local agrinom, about 5% of farmers in the area use chemical pesticides to control ants and other pests. 2 Other pests such as nematodes or fungal diseases may be responsible for similar crop damage.
119 Table 45. Agricultural p ests described by informants. Crop English Kreyl Latin Order: family Cabbage Diamondback m oth Cheni plutella Plutella xylostella Lepidoptera: Plutellidae Imported cabbage w orm Cheni chou Pieris rapae Lepidoptera: Pieridae Cabbage root m aggot Cheni rasin 1 Delia brassicae Diptera: Anthomyiidae Aphids Pichn Hemiptera: Aphididae Banana/ Plantain Banana root weevil Ti mwaygen Cosmopolites sordidus Coleoptera: Curculionidae Corn Cut worm Corn ear worm Cheni Cheni Agrotis ipsilon Helivoverpa zea Lepidoptera: Noctuidae Lepidoptera: Noctuidae Corn silk fly maggot V Diptera: Ulididae Cassava Cheni Cowpea Stinkbug Leaf-footed bugs Aphids Pinz Pinz Picho n Hemiptera: Pentatomidae Hemiptera: Coreidae Hemiptera: Aphididae Sweet Potato Sweet potato weevil Ti mwaygen Cylas formicarius Coleoptera: Curculionidae Sugarc ane Grubs Maw oka Aphids/s cale insects Pichon Tomato Caterpillars Cheni Yam Grubs likely a b eetle Maw oka Heteroligus spp Coleoptera:Curculionidae Other pests including ants and termites attack planted seeds generally consume seeds and destroy seedlings. 1 Root damage shown to Sondeo team is likely D. brassicae
120 Table 46 Insecticides used in the greater Bohoc region. Trade name Chemical Pesticide Family Signal word Use Selectivity Toxic to fish, bees, natural enemies1 Sevin Carbaryl Carbamate Warning Vegetables Broad spectrum Highly toxic Decis Deltamethrin Synthetic Pyrethroid Danger Vegetables Broad spectrum Highly toxic Match 050 DC 2 Lufenuron IGR 3 Warning Vegetables Selective Non toxic Plagafin Cypermethrin Synthetic Pyrethroid Warning Seed conservation Vegetables Broad spectrum Highly toxic Kayazol Diazinon O rganophosphate Danger Vegetables Animals Broad spectrum Highly toxic Lindagrain Lindane Organochlorine Danger Seed c onservation Broad spectrum Highly toxic Malathion Malathion O rganophosphate Caution Vegetables Animals Broad spectrum Highly toxic Neem w ater d lo ni m4 Azadirachtin Tetranortriterpenoid Caution Vegetables Pepper w ater d lo piman5 Capsican Capsaicinoid Caution Vegetables 1 Adapted from Mossler 2007. 2 The grower may have been referring to Mattch, a naturalyte pesticide made from Bacillus thurigiensis (Bt). The Bt strains in this product are selective for Lepidoptera 3 Insect Growth Regulator, Chitin synthesis inhibitor. 4,5 Biorational artisanal insecticides made by growers neem, nim, ( Azadirachta indica ) leaves and chili pepper, piman swazo, ( C. annuum ).
121 CHAPTER 5 THE IMPACT OF FOUR WEED MANAGEMENT STRATEGIES ON INVERTEBRATE POPULATIONS IN CENTRAL HAITI Introduction Weed management practices impact insect activity (Norris and Kogan 2000, Altieri 1994). Nowhere is this more important than in the tropics where subsistence farmers have limited access to technologies to manage both weed and pest populations, which are ser ious constraints to production (Akobundu 1991; Hobbs and Bellinder 2004). Hand weeding is the most common management technique in developing countries (Hobbs and Bellinder 2004). Failure to completely remove weeds is often seen as the result of labor constraints (Altieri et al.1987, Akobundu 1991). However, in many traditional agricultural systems, weeds are deliberately left in association with crops, despite the potential yield losses ( Akobundu 1991, Altieri et al. 1987). Subsistence farmers who rely on weeds as a source of animal forage, botanical pharmaceuticals, and for home consumption often practice nonclean cultivation (Altieri et al. 1987, Hillocks 1998). This relaxed weeding regime ( Altieri et al. 1987) provides weedy refugia within a vegetable cropping system that may impact invertebrate activity (Andow 1991). Incorporating weedy refugia into agricultural fields has been a successful method for increasing beneficial arthropod populations (Andow 1991, Altieri and LeTourneau 1982) and enhancing their regulatory activity (Landis et al. 2005, Showler and Greenburg 2003, Altieri 1994, Andow 1988). Furthermore, certain pest outbreaks are more likely to occur in weedfree fields than in diversified agricultural systems that retain noncrop weeds (Altieri and Whitcomb 1980; Andow 1991; Bezerra et al. 2004). Increased plant and insect diversity within the agricultural landscape may effectively
122 reduce pest associated losses, contribute to greater community stability, and minimize risk s associated with pest outbreaks (Pimental 1961). Of course, all of these benefits must be weighed against the potential of weeds to cause serious crop losses if not properly managed. Ecologically based management strategies that require few inputs and reduce labor are most accessible for low resource farmers in developing countries (Altieri 2002). Locally available mulches may provide effective weed management and reduce labor associated with weeding. Both weeds and mulches provide several ecosystem services tha t have high value, especially within the context of agricultural systems with low external inputs, including organic and subsistence agriculture. Both weeds (Gliessman 1988) and organic mulches (Powers and McSorley 2000, Ozores Hampton 1998) provide a protective soil covering that mitigates erosion, slows nutrient and water losses, and contributes to soil organic matter. Mulching offers the added benefit of reducing weed competition, which can positively impact crop yields (Ozores Hampton 1998) and may reduce labor associated with handweeding (Erenstein 2003). Also, t he use of mulches can dramatically influence the invertebrate community within the agroecosystem (Thomson and Hoffman 2007). Mulches alter the microclimate of the soil surface, increasing humidity and offering protection from temperature extremes (Riechert and Bishop 1990), all of which can dramatically influence the invertebrate community (Thomson and Hoffman 2007). Natural enemies including predatory Coleoptera (Johnson et al. 2004, Egger s and Heimbach 2001) and spiders (Riechert and Bishop 1990, Eggers and Heimbach 2001) increased with mulching as did their contribution to pest control (Stoner et. al 1996, Brust 1994, Riechert and Bishop 1990). Straw mulch reduced aphid colonization in
123 b road bean (Eggers and Heimbach 2001) and is thought to interfere with host finding in aphids (Dring et al. 2006), and thrips (Larentzaki et al. 2008). Within the context of an integrated approach to pest and weed management, it is essential for growers a nd researchers to consider the impact weeds and weed management practices may have on arthropod population dynamics (Shellhorn and Sork 1997). The goal of this study was to observe the effect of four weed management practices available to subsistence farm ers in Haitis central plateau on the soil surface invertebrate community. Materials and Methods Field experiments were conducted at the Haitian American Friendship Foundation (HAFF) (19 17N, 72 4W) located in Bohoc, Haiti, during the fall 2009 growi ng season. The experimental site had been maintained in weedy fallow for 2 years followed by a cowpea ( Vigna unguiculata) green manure planted in July 2009. Cowpea biomass was incorporated into the soil with a hoe one week prior to transplanting at the b eginning of the experiment Individual plots measured 3 m x 1 m and contained 2 rows of peppers ( Capsicum annuum L.) with 0.5 m spacing. On 29 September 2009, 6 week old transplants of local mixed pepper germplasm were transplanted by hand 0.3 m apart resulting in a density of 12 plants per plot. Four weed management treatments were replicated four times in a randomized block design. The treatments included: 100% weeded control, a partially (50%) weeded treatment, and two locally available mu lches: kodogrenn ( Paspalum sp .) hay, and composted sugarcane ( Saccharum sp .) bagasse. Bagasse is the residual fibrous bi product of sugarcane juice extraction. In the partially weeded treatment, a refuge strip of the natural weed complex was allowed to grow between the two rows of peppers, while 0.3 m2 surrounding each pepper plant
124 was kept weedfree. Mulches were applied to a thickness of 8 cm. All plots were weeded manually as needed and the time required to maintain each plot recorded. Peppers wer e fertilized at transplanting with ca. 475 mL of locally available guano. Space between plots (0.5 m) was manually cultivated to maintain bare fallow. Plants were handwatered throughout the growing season. Experimental blocks were separated by 1.5 m of unincorporated cowpea, that was allowed to go to seed. The number of pepper plants remaining in each plot at the end of the season was recorded as the stand count on 16 December. Invertebrate Sampling Invertebrate communities were sampled using pitfall traps (Triplehorn and Johnson 2005) and board traps (Cole 1946). Clear polyethylene deli containers (11 cm in diameter x 4.5 cm deep; 236 mL) (Gainesville Paper Company, Gainesville, FL) were used as pitfall traps and buried so that the upper edge was fl ush with the soil surface. The traps were filled three quarters (ca. 180 mL) with a 0.05% detergent solution to the break surface tension and prevent escape. An individual pitfall trap was randomly placed between the pepper rows near the center of the plot. Traps were left in the field for 24 h every two weeks beginning 10 d post transplanting. Board traps (Cole 1946) consisting of wood squares (15 cm x 15 cm x 2.5 cm) were randomly placed on the soil surface (under mulches), between pepper rows in each plot on 7 October. Every two weeks boards were lifted and macroinvertebrates on the underside of the board and on the soil surface covered by the board were identified to order and family when possible and recorded.
125 Statistical Analysis Data from invert ebrate sampling were analyzed by repeated measures analysis of variance (ANOVA) using the GLM procedure if SAS (SAS Institute 2008). Invertebrates collected in traps were usually identified to order and family level. In many cases, orders represented by only a few individuals in several families were grouped together by order or feeding guild for statistical analysis. Stand counts and timed weeding data were subjected to ANOVA using the SAS GLM procedure. When ANOVA was significant, means were separated with the least significant difference (LSD) test at P 0.10 (SAS Institute 2008). Results Because our experiments were initiated at the onset of Haitis dry season, weed management was less labor intensive than other times of the year. However, both the hay (0.44 0.08 min./plot) and bagasse (0.59 0.11 min.) treatments required less (F = 5.63 d. f. = 3, 41; P = 0.0025) manual weeding than the weedy (1.24 0.25 min.) and control (1.29 0.22 min.) plots. At the end of the experiment, the stand count in hay mulch (8.00 0.91 plants/plot) was lower (F = 5.03; d. f. = 3, 9; P = 0.0256) than the bagasse (11.00 0.71 plants), weed (10.50 0.29 plants) and control (11.00 0.41 plants) treatments. Greater numbers ( P = 0.0054) of Collembola were found in pitfall traps in mulched plots (hay or bagasse) than the control ( Table 51). Microhymenoptera were more abundant in the bagasse than in the hay mulch and control treat ments. Spiders and the group t otal natural enemies which included predatory Coleoptera, Diptera, Hemiptera, Hymenoptera, parasitoids, and spiders were most numerous in bagasse. More ( P = 0.0291) thrips were trapped in mulched plots than unmulched plots. Leaving weeds in
126 the plots increased (P = 0.0178) the number of auchenorrhynchans collected in pitfall traps, but this was the only instance in which invertebrate numbers in weed plots were greater than in control plots. Many groups of invertebrates wer e not affected by these treatments including: Orthoptera, Isoptera, Heteroptera, Aphididae, Coleoptera, Formicidae, Diptera, and among the noninsect invertebrates (snails, millipedes and pillbugs). Spiders were most abundant ( P = 0.0327) under board traps placed in the weed treatment ( Table 52 ), otherwise no treatment effects were observed when board traps were used. Discussion Mulches can have diverse effects on invertebrates at ground level and in the canopy (Thomson and Hoffmann 2007). The additional organic matter provided by the mulch treatments, especially the bagasse, appeared to contribute to differences among treatments. Collembola were found in greatest numbers in the mulched treatments, which invariably had more surface organic material that the control. These detrivores play an important role in decomposition (McSorley and Powers 2000), increase nutrient availability (Thomson and Hoffman 2007), and support populations of generalist predators including spiders and predatory Coleoptera (Halaj et al. 2004), especially early in the growing season (Settle et al 1996). Predator numbers may increase or decrease based on the availability of prey populations of detrivorous invertebrates like Collembola (Oelbermann et al 2008). Higher populations of spiders and natural enemies in the bagasse treatment may be related to high numbers of Collembola or other prey like thrips Aboveground, plant feeding thrips are usually assessed by other sampling methods (Southwood and Henderson 2000). However, some thrips reside in litter, feeding as fungivores or predators (Triplehorn and Johnson 2005). Like
127 Collembola, these thrips may be stimulated by organic matter and could be a source of prey for various predators Weeds provide additional complexity in terms of plant architecture and odor profiles that can influence host finding behavior and the success of pests and natural enemies alike (Capinera 2005, Cromartie 1975, Tahvanian and Root 1972). When weed refugia were incorporated into the plot s, only auchenorrhyncha were more abundant in weedy plots than in the control Also, auchenorrhyncans were the only pest organism that increased in response to treatments. During this growing season, mulches were more useful than weeds in encouraging beneficial invertebrates within the soil surface community. Board trapping and pitfall trapping were selected as sampling methods for these experiments because of local complaints of snails, ants, and problems with crop losses associated with mulch use. Thes e methods are effective for sampling the soil surface community (Triplehorn and Johnson 2005, Cole 1946). However, later in the season as mulches decomposed and rainfall became increasingly intermittent, the soil conditions became very dry, and the number of organisms observed hiding under board traps decreased. This may have been A wider range of invertebrates were captured by pitfall traps which provide information about the natural patterns of locomotion (Powell et al. 1996) and target the soil surface community. Although pitfall trapping is not the most effective sampling method for small flying insects, differences in populations of microhymenoptera were observed. Mulches and the vegetational diversity provided by weedy refugia influenced a number of arthropods typically not associated with the soil surface community as reflected in pitfall trap data.
128 Although there were no differences among the treated plots in the numbers of stem chewing invertebrates observed in traps by the end of the season, more plants were destroyed in the hay mulch treatment than the others. Most of the damage appeared to be induced by chickens which frequented t he garden. Freerange poultry such as chickens, guinea hens, and turkeys, as well as feral dogs are the sourc e of yield losses in this region of Haiti. The attractiveness of a mulch to freerange animals in a region that lacks effective fencing is a concern that growers must consider if adopting mulching as a weed management tactic. The intermittent presence of foraging chickens in the study area may have influenced the invertebrate populations and provide an additional source of error. Several groups were unaffected by the weed management tactics implemented in this vegetable system. The two mulches and weeding regimes tested did not impact the number of invertebrates including millipedes, pillbugs and snails nor nonpredatory coleopteran, heteroptra, othopteran or dipterans found in pitfall traps. It is plausible that the lack of differences was the result of small plots size. However, the areas sampled reflect the reality of rural farmers in central Haiti who grow vegetables on relatively small pieces of land or kitchen gardens that occur within the diverse background of a mixed cropping system landscape that includes weeds, fruit trees, and living borders Our results suggest that the tested weed management techniques impact certain groups within the arthropod community more than others. Composted bagasse mulch positively influenced the number of natural enemies more than the other weed management strategies tested. Mulches reduced weeding tim e throughout the season as well; however, the initial labor cost associated with acquiring and applying mulch was not within the scope of this study. Future studi es evaluating the limitations and
129 acceptance rates associated with mulch use among rural farmers is warranted. Other locally available mulches including yard waste, which is commonly burned, or cover crop residues provide both a source of organic matter and mulch for subsistance farmers; however, based on our results, Paspalum hay cannot be recommended as mulch in this system. Investigators should continue to study the merit s of bagasse mulch in other vegetable systems.
130 Table 5 1. The influence of four weed management tactics invertebrate numbers per pitfall trap. Treatment ANOVA 1 Invertebrate group Hay Bagasse Weeds Control F value P Collembola 12.10 2.47 ab 15.55 2.45 a 8.40 1.46 bc 6.20 1.15 c 4.59 0.0054 Orthoptera 0.35 0.13 0.35 0.11 0.30 0.13 0.30 0.13 0.05 0.9837 Isoptera 0.15 0.11 1.60 1.35 0.05 0.05 0.20 0.12 1.18 0.3227 Auchenorrhyncha 0.50 0.17 b 1.35 0.32 ab 2.25 0.59 a 1.15 0.35 b 3.58 0.0178 Heteroptera 0.10 0.10 0.05 0.05 0.30 0.15 0.05 0.05 1.49 0.2235 Aphididae 0.15 0.08 0.10 0.07 0.20 0.09 0.20 0.09 0.32 0.8143 Thysanoptera 1.10 0.36 a 2.65 0.65 a 1.00 0.31 b 1.05 0.40 b 3.17 0.0291 Coleoptera 0.25 0.10 0.40 0.15 0.30 0.16 0.30 0.21 0.15 0.9281 Microhymenoptera 0.35 0.15 b 0.90 0.20 a 0.50 0.14 ab 0.35 0.13 b 2.64 0.0559 Formicidae 2 16.85 1.87 25.90 9.33 31.20 10.47 42.20 15.85 1.08 0.3648 Lepidoptera 0.00 0.00 0.10 0.10 0.10 0.07 0.10 0.07 0.56 0.6449 Diptera 4.75 0.87 6.55 1.08 5.95 1.01 5.75 0.89 0.59 0.6251 Aranae 0.40 0.11 b 0.85 0.18 a 0.85 0.20 a 0.45 0.15 ab 2.46 0.0696 Diplopoda 0.45 0.24 0.15 0.11 0.20 0.12 0.10 0.07 1.15 0.3343 Isopoda 0.00 0.00 0.20 12 0.15 0.08 0.05 0.05 1.43 0.2406 Total Natural Enemies 3 1.25 0.24 b 2.45 0.30 a 1.70 0.23 b 1.65 0.31 b 3.55 0.0185 Invertebrate data are means SE of 4 replications. Means in rows followed by the same lett er do not differ according to LS D No letters in rows indicate no significant differences at P 1 Analysis of variance results; F and P values; d. f. = 3, 73. 2 Primarily Solenopsis spp. with 1 5 Odontomachus spp. 3 Total natural enemies include potential predators (Dermaptera, Heteroptera, Coleoptera, Diptera, Aranae) plus parasitoi ds (Hymenoptera), but not Formicidae.
131 Table 5 2 The influence of four weed management tactics on invertebrate numbers per board trap. Treatment ANOVA 1 Invertebrate group Hay Bagasse Weeds Control F value P Collembola 0.40 0.13 0.75 0.32 0.90 0.61 0.20 0.16 0.80 0.5003 Orthoptera 0.10 0.07 0.20 0.12 0.00 0.00 0.00 0.00 1.93 0.1329 Isoptera 0.10 0.10 0.00 0.00 0.15 0.15 5.20 4.99 1.05 0.3769 Formicidae 1.45 0.43 3.80 2.46 1.90 1.24 0.40 0.13 1.09 0.3599 Aranae 0.20 0.09 b 0.10 0.07 b 0.50 0.15 a 0.15 0.08 b 3.08 0.0327 Diplopoda 0.20 0.16 0.15 0.11 0.00 0.00 0.00 0.00 1.16 0.3323 Isopoda 0.90 0.56 0.35 0.21 0.02 0.12 0.00 0.00 1.64 0.1884 Stylommatophora 0.70 0.29 0.60 0.25 0.50 0.26 0.30 0.11 0.56 0.6412 Invertebrate data are means SE of 4 replications. Means in rows followed by the same letter do not differ according to LS D test. No letters in rows indicate no significant differences at P 1 Analysis of variance results; F and P values; d. f. = 3, 73.
132 CHAPTER 6 SURVEY OF INVERTEBRATE FAUNA OF SOIL CON SERVATION BARRIERS I N CENTRAL HAITI Introduction Haitis landscape is characterized by steep slopes and ubiquitous land degradation, the result of continuous deforestation and massive soil erosion ( Lewis and Coffey 1985). For decades, development organizations have promoted soil conservation technology including the construction of terraces, rock walls, and tree plant ing, including the establishment of contour hedgerows and alley cropping systems (Bayard et al. 2007). Erosion control barriers are established on the contour of hillsides such that over time terrac es form upslope (Shi et al. 2009). To mitigate erosion i n central Haiti, farmers use both living barriers, ( ranp vivan) also called contour hedgerows (Shi et al. 2009 ) and nonliving barriers, ( ranp m) Living barriers typically consist of either densely planted leguminous trees including Leucena spp. (Shannon et al. 2001) and Gliricidia sepium (Sridhar et al. 2001); grass species with dense root systems, like napier ( Pennisetum purpureum ) (Angima et al. 2002), guinea ( Panicum maximum ) (Shannon et al. 2001) and vetiver (Chrysopogon zizanioides ) (S ivamohan et al. 1993) ; or farmer innovations that include highvalue perennial or longlasting annual crops planted along the contour of hillsides (Murray and Bannister 2004). Nonliving barriers include permanent rock terracing and the seasonal accumulat ion of agricultural r esidues ( pai ), whic h are destroyed with plowing. Both living and nonliving erosion control barriers provide undisturbed habitat for invertebrate fauna. Living barriers offer additional vegetational resources and architecture that can provide food and refuge to beneficial and pest organisms within the agricultural landscape that are not consistently available in annual cropping systems
133 (Altieri and Nichols 2004 b ). The integration of contour hedgerows into a traditional cropping system provides a permanent habitat for invertebrate fauna that can maintain stable populations of pest and natural enemies alike (Altieri and Nichols 2004b ). Little is known about the invertebrate population on the western side of Hispa ola. Haitis politic al instability and lack of infrastructure make research difficult. The objective of this study is to survey the resident organisms exploiting the soil surface habitat behind erosion control barriers in north central Haiti Materials and Methods Study Area Surveys were conducted on farms in four communities near the intersection of National Route 3 and the Bohoc River (19 17N, 72 4 W), roughly 22 km north of Hinche, the capital of the Central Plateau. The Bohoc River marks the division between two of Haitis ten territorial divisions, the Central Plateau ( Plato Santral ) and Department of the North ( Depatman N). The communities of Bohoc, LaJeune and Terre Glisse are approximately 350 meters above sea level and are a part of the Department of the No rth, while the community of Sekadi sits among the denuded Massif du Nord mountains that flank the eastern edge of the Central Plateau. Central Haiti has bimodal rainy seasons (Butterman 1997). The early rainfall season begins in May and extends until June. A mid summer drought typically during the month of July, especially during El Nio years, is followed by a late rainfall season that spans August to November ( Butterman 1997). Agriculture is the predominant economic activity in the greater Bohoc area. Most rural Haitian farmers are the owner operators of small, fragmented landholdings of poor soil fertility (Smucker et al. 2002). The average farm size is less than 2 hectares,
134 and distributed among smaller pieces of land ( Wiens and Sobrado 1998) t hat farmers rotate ; keeping at least one parcel in fallow to improve soil fertility, graze livestock, or produce charcoal. Rural Haitian farmers practice intensive polyculture that typically include the simultaneous and relay intercropping of any of the f ollowing corn ( Z ea mays ), sorghum ( S orghum spp. ), pigeon pea ( Cajanus cajans ), manioc ( Manihot esculenta), sweet potato ( Ipomea spp.), and beans ( Phaseolus spp.) (Parafina 1993). Sites sampled in this study were characterized as either mixed cropping systems which includes any combination of the crops listed above, or fallow domin ated by colonizing grass es such as Paspalum spp. and T hemeda quadrivalvis l ocally referred to as Kodogrenn and Mada n m Michel respectively Sampling sites consisted of sm all ( <1.5 ha) parcels of land with permanent erosion control barriers that had been installed for at least 2 years Two general types of erosion control barriers were evaluated, barriers made of rock ( misek ) and contour hedgerows consisting of various agroforestry species. Sites were further characterized by land use and cropping pattern (Table 6 1). Invertebrate S ampling Clear polyethylene deli containers (11 cm in diameter x 4.5 cm deep; 236 mL) (Gainesville Paper Company, Gainesville, FL) were used as pitfall trap s (Triplehorn and Johnson 2005) and each was buried so that the upper edges were flush with soil surface. The traps were filled three quarters (ca. 180 mL) with a 0.05% detergent solution to the break surface tension and prevent escape. Traps were placed at least 12 m apart upslope from the contour of the erosion control barrier. A minimum of 5 traps were set at each sampling site for 48 hr. Contents were evaluated under a dissecting microscope and collected arthropods were identified to order and family level. In many
135 cases orders represented by only a few individuals in several families were grouped together by suborder or feeding guild. Results Both living (Table 6 2 ) and nonliving (Table 6 3 ) soil conservation barriers in this survey were characterized by similar arthropod groups especially Collembola, ants, and spi ders Fewer flying insects (including microhymenopterans, Diptera and Lepidoptera) were found in pitfall traps al though these groups are usually sampled using other techniques. The majority of the ants in pitfall traps were Solenopsis spp. ; h owever up to 30% of the ants in each trap were Odontomachus spp. B oth of these ants can serve as important predators (Fisher and Cover 2007). Near the rock barriers (Table 6 3 ) Collembola were three times as abundant in the mixed cropping system, when compared to the fallow environment, which consists primarily of grasses and woody shrubs. The only erosion control barrier t hat provided habitat for Stylommatophora ( snails ) was the napier grass, P. purpureum Additionally, Dermaptera were most often found in traps set in P. purpureum barriers Discussion As topsoil and organic matter accumulate upslope from a soil conservat ion barrier a biologically rich habitat develops that can support a diverse community of organisms at the soil surface. S oil decomposers and nutrient recycling fauna including ants, Collembola, thrips, termites, and mites were captured in pi t fall traps placed upslope from living and nonliving barriers. However, living barriers may further contribute soil amelioration through leaf litter accumulation and prunings N fixation, and nutrient recycling Contour hedgerows increase the complexity of the agro ecosystem by providing plant resources, alternate prey, architecture and habitat required by
136 arthropods which is especially important during periods when the land is disturbed through cultivation or harvest ( MacLean et al. 2003, Altieri and Nichols 2004 b ) The type of trees and shrubs planted along a contour hedgerow may impact the arthropods present ( Girma et al. 2000, Mac Lean et al. 2003). Some species including G. sepium may have a repellent effect (Mac Lean et al. 2003) while others may serve as an attractant. Pennisetum purpureum is a common fodder species in the tropics that functions as an effectiv e erosion control barrier (Angima et al. 2002) In pushpull maize cropping systems in Africa, P. purpureum field borders serve as a trapcrop for stemboring insects (Khan et al. 2000). In the current study, similar insects were found in P. purpureum and other contour hedgerow crop species. However, in this survey, snails were only found in the P. purpureum barrier. Since local farmers report that snails are a major vegetable pest in this region (HansPetersen Chapter 4) further investigation is warranted. The use of contour hedgerows represents a more efficient use of marginal land that can increase overall farm production by providing a source of income, food or fodder for the farmer (Shannon et al. 2001). Inspite of demonstrated benefits ( Shannon et al. 2001, Bayard et al. 2007), the adoptio n of contour hedgerow technology in Haiti has been low and hedgerows have been abandoned throughout Ha iti (Bayard et al. 2007, Shannon et al. 2001). Although bioterraces are much more cost efficient and stable than engineered terraces (Shi et al. 2009), they must be maintained to prevent shading effects in neighboring crops, and are more susceptible to g razing animals, thievery, and relatives collecting firewood (Murray and Bannister 2004). The construction misek is both expensive and time consuming (Mutegi et al. 2008, Bayard et al. 2006). In fact, the rock borders that we evaluated were not built by the
137 landowners alone, rather their construction was coordinated and financed by nongovernmental organizations (NGOs). The calc areous hills bordering the Central Plateau supplied copius amounts of rock for building, while NGO s provided training and compensation (food for work) for a large crew of local laborers In addition to having erosion control barriers installed on their land, farmers and laborers working in this rugged terrain benefit by participating in the construction of misek Some motivated farmers have begun building soil conservation barriers as a result of exposure to this technology. Many of these farmers have also taken an interest in insects as a result of this surv ey. Over 15 farmers volunteered their land and time to participate in this survey, and many took advantage of the opportunity to observe arthropods from their fields under a microscope. Because there is little knowledge of the different roles of various invertebrates within the agroecosystem (HansPetersen Chapter 4) farmers were very interested in learning that many of the invertebrates found at the soil surface contributed to decomposition and nutrient recycling, and that natural enemies like parasitoids and predators were present in their fields. This survey was conducted during the dry season when crops like pigeon pea, sorghum and bean were nearly ready to be harvested. Most of the corn was dry on the stalk except where a second season crop was plant ed in Sekadi. The advent of the rainy season in May/June will likely change arthropod dynamics as rainfall triggers new and vulnerable growth. A survey during the rainy season might show more distinctive trends especially for univoltine species that hibernate during the dry season. Peasant cropping systems in Haiti are characterized by diversity a strategy that expands the duration of harvest and the availability of food resources for the household
138 (Smucker et al. 2002). Often the biodiversity associated with these traditional agroecosystems is comparable to those of natural systems (Altieri 1999). Thus it is not surprising that similar invertebrates were captured in various fields despite very different erosion barriers. The significance of the refugia provided by these barriers cannot be generalized, but depends on arthropods within the system (Girma et al 2000) and will require further investigation. One of the greatest strengths of the survey was that it was conducted in farmer s fields and resul ted in the generation of interest in beneficial organisms associated with erosion control barriers.
139 Table 6 1. Survey site information. Site Date Barrier t ype Community Cropping pattern Land use No. traps S 1 20 Nov Rock Sekadi Fallow Charcoal 6 B 1 14 Nov Rock Bohoc Fallow Pasture 8 S 2 20 Nov Rock Sekadi Mixed Household 8 B 3 10 Dec Rock Forestry species Bohoc Mixed Propagation 1 5 S 3 02 Dec Gliricidia sepium Sekadi Mixed Household 5 T 1 12 Dec Leucena spp. Terre Glise Fallow Pasture 9 L 1 16 Nov Saccharum sp Musa spp. Dioscorea spp. LaJeun e Mixed Household 6 S 4 02 Dec G. sepium Sekadi Fallow Pasture 7 B 2 12 Nov Pennistum purpureum Bohoc Fallow Pature 6 1 NGO farm amplifying Cajanus cajans and Manihot esculenta propagation materials
140 Table 6 2. Invertebrate numbers per pitfall trap placed in living erosion control borders ( ranp vivan) Invertebrate group S 3 (mixed) L 1 (mixed) T 1 (fallow) S 4 (fallow) B 2 (fallow) Collembola 4.80 0.92 1 10.00 3.10 60.11 16.29 9.57 1.21 21.50 5.95 Orthoptera 4.40 1.69 0.20 0.18 0.44 0.18 0.86 0.14 0.67 0.42 Dermaptera 0.00 0.00 0.60 0.36 0.00 0.00 0.00 0.00 1.00 0.82 Isoptera 0.00 0.00 0.20 0.18 0.78 0.43 0.00 0.00 2.67 1.18 Blattodea 1.00 0.55 0.00 0.00 0.56 0.34 0.29 0.19 0.67 0.33 Auchenorrhyncha 0.40 0.25 1.40 0.36 6.44 0.91 1.43 0.57 0.67 0.33 Aphididae 0.00 0.00 0.00 0.00 0.11 0.11 0.00 0.00 0.33 0.21 Hemiptera 0.40 0.40 0.40 0.22 0.56 1.13 0.29 0.19 0.17 0.17 Thysanoptera 0.20 0.20 0.00 0.00 2.78 1.08 0.86 0.34 0.00 0.00 Psocoptera 0.00 0.00 0.40 0.22 0.44 0.18 0.14 0.14 0.17 0.17 Coleoptera 0.80 0.20 0.00 0.00 1.11 0.65 0.57 0.20 1.17 0.40 Hymenoptera 2.20 0.80 2.60 0.98 10.22 5.76 0.14 0.14 4.33 0.62 Formicidae 21.60 6.39 79.80 25.28 77.89 26.46 2.43 0.65 37.33 15.82 Lepidoptera 0.20 0.20 0.20 0.18 1.56 0.65 0.71 0.36 0.50 0.23 Diptera 2.00 0.45 4.60 0.85 20.56 4.14 2.29 0.99 18.50 3.34 Acari 0.20 0.20 4.80 2.26 5.56 1.63 0.86 0.59 0.33 0.34 Aranae 1.80 0.73 1.60 1.82 2.44 0.58 1.00 0.43 2.50 0.43 Stylommatophora 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2.67 1.89 Total Natural Enemies 2 4.80 1.50 5.00 1.71 13.11 5.58 3.71 0.61 9.00 0.93 1Data are means SE based on number of traps in Table 6 1. 2Includes Hymenopteran parasitoids, predatory Dermaptera, Hemiptera, Coleoptera, Diptera, and Aranae; excludes ants.
141 Table 6 3. Invertebrate numbers per pitfall trap placed in rock erosion control borders ( m isek) Invertebrate group S 2 (mixed) B 3 (mixed) B 1 (fallow) S 1 (fallow) Collembola 21.00 3.17 1 28.00 6.35 7.00 1.55 7.17 1.82 Orthoptera 1.75 0.53 0.00 0.00 2.00 0.56 0.83 0.31 Dermaptera 0.50 0.27 0.00 0.00 0.00 0.00 0.17 0.17 Isoptera 0.38 0.18 0.00 0.00 0.13 0.12 0.17 0.17 Blattodea 0.25 0.16 0.20 0.20 0.00 0.00 0.17 0.17 Auchenorrhyncha 1.50 0.32 2.20 0.97 1.13 0.44 1.67 0.85 Aphididae 0.25 0.25 0.00 0.00 0.13 0.12 0.17 0.17 Heteroptera 0.25 0.16 1.20 0.20 0.00 0.00 0.17 0.17 Thysanoptera 0.63 0.26 4.40 0.40 0.25 0.17 0.33 0.21 Psocoptera 0.00 0.00 0.80 0.38 0.63 0.18 0.17 0.17 Coleoptera 1.00 0.33 0.40 0.25 0.13 0.12 0.50 0.34 Hymenoptera 3.38 0.68 4.20 0.58 2.75 0.94 3.00 0.97 Formicidae 104.25 26.62 5.00 1.41 41.38 13.46 53.67 14.88 Lepidoptera 0.63 0.33 0.60 0.25 0.25 0.16 0.33 0.33 Diptera 5.13 1.57 16.20 4.45 12.63 4.78 7.33 2.34 Acari 3.13 1.45 0.80 0.38 0.63 0.50 3.00 0.63 Aranae 2.63 0.60 2.80 0.38 1.38 0.53 6.67 0.92 Stylommatophora 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Total Natural Enemies 2 6.88 1.00 8.00 0.95 4.13 1.10 10.00 1.16 1Data are means SE based on number of traps in Table 6 1. 2Includes Hymenopteran parasitoids, predatory Dermaptera, Hemiptera, Coleoptera, Diptera, and Aranae; excludes ants.
142 CHAPTER 7 CONCLUSIONS Increasing habitat complexity in agricultural fields can reduce pest colonization and increase natural enemy populations (Altieri and Letourneau 1982). Theories regarding the mechanisms behind this phenomenon have been extensively reviewed (Landis et al. 2000, Smith and McSorley 2000, Russell 1989, Sheehan 1986). In the present study, the influence of habitat management on the invertebrate community was evaluated in multiple experiments that integrated noncrop vegetation as border crops, altered non crop plant diversity through weeding regimes and mulch application and examined permanent soil conservation structures. Pest and natural enemy abundance, weed richness and biomass, and yields were compared in pepper ( Capsicum annuum L.) plots with different w eed densities: 100% weedfree; 50% weedfree; 50% weedfree intercropped with bush bean ( Phaseolus vulgaris L.) ; and an unweeded control. Pan traps, pitfall traps, sticky card s, and in situ counts were used to estimate the relative abundance of several arthropod groups. Insect numbers in the intermediate (50% weeded) treatments were typically not distinctly different from the other treatments; however several differences were detected when the 100% weedfree treatment was contrasted with the unweeded control. Beneficial insects were generally found in greater numbers in the unweeded treatment than in weedfree plots. The number of different arthropod taxa present correlated with weed richness. An exception to these trends occurred with flea beetles ( Altica spp.), which were most abundant in weedfree plots. Despite increases in arthropod richness, leaving weeds within the plot did not result in increased biological control and negatively affected pepper yields except in the bean intercrop treatment.
143 The influence of intercropping strips of noncrop vegetation on the aboveground arthropod community was assessed, including increase of natural enemy populations and interference with pest colonization in an adjacent squash crop. Four noncrop plot border treatments included: sorghum x sudangrass hybrid ( Sorghum bi color [L.] Moench x S. sudanense [Piper] Stapf); pigeon pea, ( Cajanus caj an [L.] Millsp.); the native weed complex, and a bare ground control. These border treatments were selected because of their acceptability and usefulness among rural Haitians as food, and forage. Border crops were established on both sides of experimental plots containing yellow crookneck squash, ( Cucurbita pepo L.). Sticky cards, pitfall traps, pan traps and in situ counts were used to assess differences in the arthropod community within each of the border treatments and the adjacent squash crop. Natural enemies were most abundant in the native weed complex and pigeon pea borders; however the spillover of natural enemies into the neighboring crop was only observed in 2008 when pr edatory Coleoptera were most abundant in both the sorghum sudangrass treatm ent and adjacent squash. Border crops did not influence the mo vement of thrips and whiteflies; however, in situ aphid counts were lower on squash bordered by sorghum sudangrass than in the control. None of the border treatments could prevent a heavy infestation of melonworm ( Diaphania hyalinata L.), which defoliated and killed many of the squash plants. Additionally, c onversational interviews were conducted with peasant farmers in north central Haiti to assess pest management practices. The S ondeo method was an effective technique for gathering copius information in a short amount of time. The majority of the agricultural land in these communities is planted wi th a traditional crop mix or sugarcane, both of which are rainfed and receive no external inputs. Several
144 nonchemical pest management techniques are integrated into traditional farming systems including crop rotation, intercropping, removing diseased plant residues, and adjusting planting dates to avoid pest and disease pressure. F armers in this region have access to agrochemicals at local markets and agricultural supply stores in neighboring towns. A variety of pesticides ranging from moderately to highly toxic can be purchased without label information, safety instructions, or often without appropriate packaging. The majority of participants had only used pesticides to conserve grain, with the exception of vegetable growers who employ pesticides regul arly to control cabbage pests. Very few interviewees had received any pesticide training and none of the Sondeo participants were familiar with beneficial insects. Integrated p est m anagement (IPM) training would fill a necessary information gap for farmers in these communities, especially among those with irrigated vegetable plots. Training efforts should focus on existing community groups and young people. The effect of weed management practices on the soil surface invertebrate community was further evaluated in cultivated pepper in a field experiment in central Haiti. The treatments were: 100% weeded control, 50% of the plot area weeded, and two locally available mulches: composted sugarcane ( Saccharum spp.) bagasse mulch and hay ( Paspalum spp.) mulc h. Although many groups of invertebrates were unaffected, the use of bagasse resulted in highest numbers of Collembola, microhymenoptera, and spiders. Leaving weeds in plots increased auchenorrhynchans, but did not affect any other invertebrates. Mulchi ng with bagasse was more useful than maintaining weed refuge for increasing numbers of beneficial arthropods in the soil surface community. The use of Paspalum spp. mulch resulted in lower stand counts of pepper and is not recommended as a mulch in this s ystem.
145 The addition of permanent erosion control structures within the agricultural landscape provides uncultivated perennial refugia that can serve as invertebrate habitat. A survey of selected invertebrate groups associated with different soil conserv ation barriers in central Haiti was conducted during November and December of 2009. Pitfall traps were used to characterize the invertebrate communities within living and nonliving erosion barriers in both successional and mixed cropping systems. Invert ebrates from 17 orders were collected. Ants, Collembola, and spiders were most closely associated with the barrier habitat regardless of the type of border or cropping system. Farmers enthusiastically participated in data collection and expressed interes t in learning more about the role of captured invertebrates in their fields. Much of the literature concerning habitat manipulation in agricultural systems is focused on the response of a particular pest or natural enemy guild to increased complexity. Ho wever, any action that alters the agroecosystem may affect members of the invertebrate community at multiple trophic levels not directly impacting the crop. In the present studies, several habitat manipulations including the addition of noncrop vegetation, border crops, mulches, and erosion control barriers, were incorporated into various agroecosystems. The impact of these actions on the entire invertebrate community was assessed in a pesticidefree environment within vegetationally diverse agricultural landscapes Therefore, the results cannot be generalized to conventional vegetable production systems that operate against a less varied background. Both the lack of agrochemicals and high levels of plant diversity at the field and landscape levels like ly provided a stabilizing buffer that prevented serious pest outbreak in pepper experiments (Pimental 1981, Altieri and Nichols 2004) However, in the squash study, h abitat manipulation was not a sufficient measure of crop protection to deter
146 melonworm D iaphania hyalinata Rather it is one of many tools that farmers can employ in concert with other IPM practices to sustainably manage pests.
147 LIST OF REFERENCES Akobundu, O. I. 1991. Weeds in human affairs in subsaharan Africa: Implications for sustainable food production. Weed Tech. 5: 680690. Altieri, M. A. 1994. Biodiversity and pest management in agroecosystems. Food Products Press Inc. New York, NY. Altieri, M. A. 1999. The ecological role of biodiversity in agroecosystems. Agric. Ecosys. Environ. 74: 1931. Altieri, M. A 2002. Agroecology: the science of natural resource management for poor farmers in marginal environments. Agric. Ecosys. Environ. 93: 124. Altieri, M. A., and W. H. Whitcomb. 1980. Weed manipulation. J. Environ. Manag. 4: 438489. Altieri, M. A., and D. K. Letournea u. 1982. Vegetation management and biological control in agroecosystems. Crop Prot. 1: 405430. Altieri, M. A., M. K. Anderson, and L. C. Merrick. 1987. Peasant agriculture and the conservation of crop and wild plant resources. Conserv. Biol. 1: 4958. Altieri, M. A., and C. I. Nichols. 2004a Biodiversity and pest management in agroecosystems 2nd Edition. Food Products Press, Binghamton, NY. Altieri, M. A., and C. I. Nicholls. 2004b. Ef fects of agroforestry systems on the ecology and management of insect pest populations, pp.142154. In G. M. Gurr, S .D., Wratten, and M. A. Altieri ( eds. ) E cological engineering for pest management: A dvances in habitat manipulation for arthropods. Comstock Pub. Asso ciates/Cornell University Press, Ith aca, NY. Andow, D. A. 1988. Management of weeds for insect manipulation in agroecosystems, pp. 265301. In M. A. Altieri and M. Z. Liebman ( eds .), Weed management in agroecosystems: E cological approaches. CRC Press, Boca Raton, FL. Andow, D. A. 1991. Vegetational diversity and arthropod population response. Ann. Rev. Entomol. 36: 561586. Angima, S. D., D. E. Stott, M. K. ONeill, C. K. Ong, and G. A. Weesies. 2002. Use of Calliandra n apier grass contour hedges to control erosion in central Kenya. Agric. Ecosys. Environ. 91: 1523. Bach, C. E. 1984. Plant spatial pattern and herbivore population dynamics: P lant factors affecting the movement patterns of a tropical cucurbit specialist ( Acalymma innubum ). Ecology 65: 175190.
148 Brberi, P. 2002. Weed management in organic agriculture: Are we addressing the right issues? Weed Res. 42: 177193. Basu, A. N. 1995. Bemisia tabaci (Gennadius) crop pest and principal whitefly vector of plant viruses. Westview Press Boulder, CO Bayard, B., C. M. Jolly, and D. A. Shannon. 2006. The adoption and management of soil conservation practices in Haiti: The case of rock walls. Agricul. Econ. Rev. 7: 2839. Bayard, B., C. M. Jolly, and D. A. Shannon. 2007. The economics of adoption and management of alley cropping in Haiti. J. Environ. Manag. 84: 6270. Bentley, J. W. 2003. The mothers, fathers and midwives of invention: Zamoranos natural pest control course. In G. Stoll (ed), Natural Crop Protection in the Tropics. Margraf Publishers, The Netherlands. http://www.naturalcropprotection.margraf verlag.de/mothers.htm Bezerra, M. S., M. R. V. de Oliveira, and S. D. Vasconcelos. 2004. Does the presence of weeds affect Bemisia tabaci (Gennadius) (Hemiptera: Aleyrodidae) infestation on tomato plants in a semi arid agroecosystem? Neotrop. Entomol. 33: 769775. Bishop, C 1995. Nouvo gid kilti legim. Mission Baptiste dHaiti, Kenscoff, Haiti. Blythe, K. 2002. Livelihoods and sustainability in La Visite National Park, Haiti. MS thesis, University of Florida, Gainesville. Bonte, D., M. Lukac, and L. Lens. 2008. Sta rvation affects predispersal behaviour of Erigone spiders. Bas. Appl. Ecol. 9: 308315. Breuer N. E., S. Adhikari, R. BrownSalazar, J. A. Clavijo, H. N. HansPetersen, N. C. Kawa, R. Patarasuk, and P. E. Hildebrand. 2008. Extension agent perspectives of climate forecasts, and the AgClimate decision support system. Technical Report Series: SECC 0801, Southeast Climate Consortium University of Florida, Gainesville, FL. http://seclimate.org Brust, G. E. 1994. Natural enemies in straw mulch reduce Colorado potato beetle populations and damage in potato. Biol. Control. 4: 163 169. Bugg, R. L., L. E. Ehler, and L. T. Wilson. 1987. Effect of common knotweed ( Polygonum aviculare ) on abundance and efficiency of insect predators of crop pests. Hilgardia 55: 151. Butterman, H. C. 1997. Rainfall variability in Haiti and its relation to global and regional scale climate indices. Ed.D. dissertation, Teachers College, Columbia University, New York.
149 Buss, T., and A. Gardner. 2006. Why foreign aid to Haiti failed. National Academy of Public Administration, Washington, DC. Capinera, J. L. 2005. Relationships between insect pests and weeds: An evolutionary perspective. Weed Sci. 53: 892901. Chase, C. A., and O. S. Mbuya. 2008. Greater interference from living mulches than weeds in organic broccoli production. Weed Technol. 22: 280285. Chen, K. H. and G. F. Murray. 1976. Truths and untruths in village Haiti : A n experiment in third world survey research, pp. 241262. In J. F. Marshall and P. Polgar (e ds. ) Culture, natality, and family planning. Carolina Population Center, University of North Carolina, Chapel Hill, NC. Christie, R. G., and J. R. Edwardson. 1994. Light and electron microscopy of plant virus inclusions. Monograph 9, revised. Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL. (CIA ) Central Intelligence Agency. 2009. The w orld f actbook Washington, DC: https://www.cia.gov/library/publications/theworld factbook/index.html Cole, L. C. 1946. A study of the cryptozoa of an Illinois woodland. Ecol. Monogr. 16: 4986. Coll inson, M. 2001. Institutional and professional obstacles to a more effective research process for smallholder agriculture. Agric. Syst. 69: 2736. Compton, S. G. 2002. Sailing with the wind: D ispersal by small flying insects, pp. 113 133. In J. M. B ullock, R. E. Kenward and R. S. Hails (eds.), Dispersal ecology. Blackwell Publishing, Malden, MA. Correia, M. 1998. Gender and poverty in Haiti technical paper 2. In Haiti: The challenges of poverty reduction. Volume II technical papers Report No. 17242 HA. Poverty Reduction and Economic Management Unit and Caribbean Country Management Unit, Latin American and the Caribbean Region. World Bank, Washington, D.C. Cromartie Jr., W. J. 1975. The effect of stand size and vegetational bac kground on the colonization of cruciferous plants by herbivorous insects. J. Appl. Ecol. 12: 517533. Damicone, J. P., J. V. Edelson, J. L. Sherwood, L. D. Myers, and J. E. Motes. 2006. Effects of border crops and intercrops on control of c ucurbit virus diseases. Plant Dis 91: 506516. Denys, C. and T. Tscharntke. 2002. Plant insect communities and predator prey ratios in field margin strips, adjacent crop fields, and fallows. Oecologia 130: 315324.
150 Dosdall, L. M., M. G. Dolinski, and P. M. Conwa 1999. The effect of tillage regime, row spacing, and seeding rate on feeding damage by flea beetles, Phyllotreta spp. (Coleoptera: Chrysomelidae), in canola in central Alberta, Canada. Crop Prot. 18: 217224. Drinkwater, L .E, D. K. Letourneau, F. Worknehvan A. H. C., Bruggen, and C. Shennan. 1995. Fundamental differences between conventional and organic tomato agroecosystems in California. Ecol. Appl. 5: 10981112. Dring, T. F., U. Heimbach, T. Thieme, and H. Saucke. 2006. Reduzierter Blattlausbefall durch Strohmulch in kologisch angebauten Kartoffeln. Mitt. Dtsch. Ges Allg. Angew. Entomol. 15: 123126. Eggers, C., and U. Heimbach. 2001. Reduction of aphid numbers by using mulching techniques in broad beans. Mitt. Dtsch. Ges Allg. Angew. Entomol. 13:253256. Ehlers, J. D., and A. E. Hall 1997. Co w pea ( Vigna unguiculata L. Walp.). Field Crops Res. 53: 187204. Eisler, R. 1986. Diazinon hazards to fish, wildlife, and invertebrates: A synoptic review. Biological Report : 85(1.9). U.S. Fish and Wildlife Service, Laurel, MD Erenstien, O. 2003. Smallholder conservation farming in the tropics and subtropics: a guide to the development and disseminat ion of mulching with crop residues and cover crops. Agric. Ecosys. Environ. 100: 1737. (Extoxnet) Extension toxicology network. 1996. Extension Toxicology Network pesticide information profiles: Diazinon. Dept. Environmental Toxicology, University of California, Davis, CA. http://extoxnet.orst.edu/pips/diazinon.htm FAO (Food and Agriculture Organization of the United Nations). 2007. Good practices for hazard risk management in agriculture. Summary report Haiti: Project phase I TCP/RLA/3101 FAO, Rome. ftp://ftp.fao.org/docrep/fao/010/ai501e/ai501e01.pdf (FAO) Food and Agriculture Organization. 2001. Conservation agricult ure: Case studies in Latin America and Africa. FAO Soils Bulletin 78. Food and Agricultur e Organization. Rome, Italy. (FAWN) Florida Automated Weather Network. 2009. IFAS, University of Florida, Gainesville, FL. http://fawn.ifas.ufl.edu Fereres, A. 2000. Barrier crops as a cultural control measure of nonpersistently transmitted aphidborne viruses. Virus Res 71: 221231. Finn, E. 2003. Developing integrated pest management techniques for managing key insect pests of blueberries in the southeastern United States. MS thesis, University of Florida, Gainesville FL.
151 Fisher, B. L., and S. P. Cover 2007. Ants of North America: a guide to the genera. University of California Press, Berkeley, CA. Flint, M. L., and P. Gouveia. 2001. IPM in practice: Principles and methods of integrated pest management. Publication 3418. University of California Statewide Integrated Pest Management Project: Agriculture and Natural Resources. Regents of the University of California Division of Agriculture and Natural Resources, Oakland, CA. Freeman, B. C. and J. Laguerre. 2002. Haitian English dictionary 4th edition. Press Evanglique, Port auPr ince, Haiti. Geie r, P. W. 1966. Management of insect pests. Annu. Rev. Entomol. 11: 471490. Gilreath, P. 200 6 Cultural controls: W indbreaks pp.184185. In J. L. Gillett, H. N. HansPetersen, N. C. Leppla, and D. D. Thomas (eds.), Growers IPM guide for F lorida tomato and pepper production. UF/IFAS Cooperative Extension, University of Florida, Gainesville, FL. Girma, H., M. R. Rao, and S. Sithanantham. 2000. Insect pests and beneficial arthropods populations under different hedgerow in intercropping systems in semiarid Kenya. Agroforest. Syst. 50: 279292. Gliessman, S. R. 1988. Ecology and management of weeds in traditional agroecosystems, pp. 237243. In M. A. Altieri and M. Liebman (eds.), Weed management in agroecosystems: E cological approaches. CRC Press, Boca Raton, FL. Goldmann, L., B. Murray, B. Wahlstrm, and J. Pronczuk. 2004. Childhood pesticide poisoning information for advocacy and acti on. Chemicals Programme of the United Nations Environment Programme (UNEP), Chtelaine, Switzerland. Gonsalves, D. 1994. Papaya diseases caused by viruses, pp 67 68. In R. C. Ploetz, G. A. Zentmyer, W. T., Nishijima, K. G. Rohrbach, and H. D. Ohr (ed s.), Compendium of tropical fruit diseases. APS Press, St. Paul, MN. Halaj, J., A. B. Cady, and G. W. Uetz. 2000. Modular habitat refugia enhance generalist predators and lower plant damage in soybeans. Environ. Entomol. 29: 383393. Hildebrand, P.E 1981. Combining disciplines in rapid appraisal: The Sondeo approach. Agric. Admin. 8: 423432. Hildebrand, P., A. J. Sullivan, N. E. Breuer, and V. E. Cabrera. 2001. Modeling diverse livelihood strategies in small farm livelihood systems using ethnographic linear programming. Food and Resource Economics Department, IFAS, University of Florida, Gainesville, FL.
152 Hildebrand, P.E., and A.J. Sullivan. 2001. Small farm livelihood systems and food security: addressing diversity. Staff Paper Series SP027. Food and Resources Economics Department, IFAS, University of Florida, Gainesville, FL. Hillocks, R. J. 1998. The potential benefits of weeds with reference to small holder agriculture in Africa. Integr. Pest Manag. Rev. 3: 155167. Hobbs, P. R., and R. R. Bellinder. 2004. Weed Management in less developed countries, pp. 12951298. In R. M. Goodman (ed.), Encyclopedia of plant and crop science. Marcel Dekker Inc., New York, NY. Hooks, C. R. R. and M. Johnson. 2003. Impact of agricultural diversification on the insect community of cruciferous crops. Crop Prot. 22:223238. Jadotte, E. 2006. Income distribution and poverty in the republic of Haiti. WP: 200613. Poverty and Economic Policy Research Network, Qubec City, QB. Johnson, J. M J. A HoughGoldstein, and M. J. Vangessel. 2004. Effects of straw mulch on pest insects, predators, and weeds in watermelons and potatoes. Environ. Entomol. 33: 6: 16321643. Kareiva, P. 1985. Finding and losing host plants by Phyllotreta : patch size and surrounding habitat. Ecology 66: 18091816. Kemp, J. C. and G. W. Barrett. 1989. Spatial patterning: impact of uncultivated corridors on arthropod populations within soybean agroecosystems. Ecology 70: 114128. Kitch, L. W., O. Boukar, C. Endodo, and L. L. Murdock. 1998. Farmer acceptability criteria in breeding cowpea. Expl. Agric. 34: 475486. Khan, Z. R., J. A. Pickett, J. van den Berg, L. J. Wadhams, and C. M. Woodcock. 2000. Exploiting chemical ecology and species diversity: stem borer and striga control for maize and sorghum in Africa. Pest Manag. Sci. 56: 957962. Landis, D. A and F. D. Menalled 1998. Ecological considerations in the conservation of effective parasitoid communities in agriculture systems, pp. 101121, In P. Barbosa (ed.), Conservation Biological Control. Academic Press, San Diego, CA. Landis, D. A., F. D. Menalled, A. C. Costamagna, and T. K. Wilkinson. 2005. Manipulating plant resources to enhance beneficial arthropods in agricultural landscapes. Weed Sci. 53: 902908. Landis, D.A., S.D. Wratten, and G. M. Gurr. 2000. Habitat management to conserve n atural enemies of arthropod pests in agriculture. Annu. Rev. Entomol. 45: 145201.
153 Lanini, W. T., and M. LeStrange. 1994. Weed control economics in bell pepper ( Capsicum annuum ) with napropamide and hand weeding. Weed Technol. 8: 530535. Larentzaki E., J. Plate, B. A. Nault, and A. M. Shelton. 2008. Impact of Straw Mulch on Populations of Onion Thrips (Thysanoptera: Thripidae) in Onion. J. Econ. Entomol. 101: 13171324. Letourneau, D. K. 1987. The enemies hypothesis: tritrophic interactions and vegetational diversity in tropical agroecosystems. Ecology 68: 16161622. Lewis, L. A. and W. J. Coffey. 1985. T he continuing deforestation of Haiti. Ambio : 14: 158160. Lewis, T. 1997. Flight and dispersal, pp. 175 196. In T. Lewis (e d. ) T hrips as crop p ests. CAB International New York, NY. Lewis, T. 1969. The distribution of flying insects near a low hedgerow. J. Appl. Ecol. 6: 443452. Lewis, W. J., J. C. van Lenteren, S. C. Phatak, and J. H. Tumlinson, III. 1997. A total system approach to sustainable pest management. Proc. Natl. Acad. Sci. 94: 1224312248. Litow, P.A. 2000. Food security and household livelihood strategies in the M aya B iosphere R eserve: the importance of milpa in the community of Uaxactn, Guatemala. M.S. thesis, University of Florida, Gainesville, FL. Luckmann, W. H., and R. L. Metcalf. 1994. The pest management concept, pp. 134. In R. L. Metcalf and W. H. Luckmann (eds.), Introduction to insect pest management, third edition. John Wiley & Sons Inc., New York, NY. Lundahl, Matts. 1984 The Roots of Haitian Underdevelopment, pp. 181203, In C. R. Foster and A. Valdman [eds.], Haiti Today and Tomorrow. University Press of Amer ica, New York, NY. MacLean, R. H., J. A. Litsinger, K. Moody, A. K. Watson, and E. M. Libetario. 2003. Impact of Gliricidia sepium and Cassia spectabilis hedgerows on weeds and insect pests of upland rice. Agric. Ecosys. Environ 94: 275288. Marino, P. C., and D. A. Landis. 1996. Effect of landscape structure on parasitoid diversity and parasitism in agroecosystems. Ecol. Appl. 6: 276 284. Marshall, E. J. P., V. K. Brown, N. D. Boatman, P. J. W. Lutman, G. R. Squire, and L. K. Ward. 2002. The role of weeds in supporting biological diversity within crop fields. Weed Res. 43: 7789.
154 Maxwell, B. D., and J. T. ODonovan. 2007. Under standing weed crop interactions to manage weed problems, pp. 1734. In M. K. Upadhyaya and R. E. Blackshaw (eds.), Nonchemical weed management: Principles, concepts and technology. CAB International Wallingford, UK. Meir, C. 1999. Improving women's participatio n in pest management training: a pilot study in Honduras, pp. 7178. In E. Van de Fliert, and J. Proost (eds.), Women and IPM: Crop protection practices and strategies. Royal Tropical Institute and Intermediate Technology Pulications, The Net herlands. Mossler, M. 2006. Beneficial agents interactions with pesticides, pp.88. In J. L. Gillett, H. N. HansPetersen, N. C. Leppla, and D. D. Thomas (eds.), Grower s IPM g uide for Florida tomato and pepper production. UF/IFAS Cooperative Extension, University of Florida, Gainesville, FL. Mossler, M., M. J. Aerts, and O. N. Nesheim 2006. Florida crop/pest management profiles: Bell peppers. Florida Cooperative Extension ServiceCir. 1240/PIO40, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL. http://edis.ifas.ufl.edu/PI040. Murphy, J. F., J. Mosjidis, M. D. Eubanks, and J. Masiri. 2008. Inter row soil cover to reduce incidence of aphidborne viruses in pumpkin. Int J. V eg. Sci. 14: 290303. Murray, D. L. 1994. Cultivating crisis: The human cost of pesticides in Latin America. University of Texas Press, Austin, TX. Murray, G. F., and M. E. Bannister. 2004. Peasants, agroforesters, and anthropogists: A 20 year venture in incomegenerating trees and hedgerows in Haiti. Agroforest. Syst. 61: 383397. Mutegi, J. K., D. N. Mugeni, L. V. Verchot, and J. B. Kungu. 2008. Combining napier grass with leguminous shrubs in contour hedgerows controls soil erosion without competing with crops. Agroforest. Syst. 74: 3749. Nentwig, W. 1988. Augmentation of beneficial arthropods by stripmanagement. Oecologia 76: 597606. Nielsen, R. 1996. Incomplete kernel set in corn. Purdue Pest Management & Crop Production Newsletter Dept. Agronomy, Purdue University, West Lafayette, IN. http://www.agry.purdue.edu/ext/corn/news/articles.96/p&c9642.htm Nyoike, T. W., and O. E. Liburd. 2010. Effect of living (buckwheat) and UV reflective mulches with and without imidacloprid on whiteflies, aphids and marketable yields of zucchini squash. Int J. Pest Manag. 56: 3139. Norris, R. F. 2005. Ecological bases of interactions between weeds and organisms in other pest categories. Weed Sci. 53: 909913.
155 Norris, R. F., and M. Kogan. 2000. Interactions between weeds, arthropod pests and their natural enemies in managed ecosystems. Weed Sci. 48: 94158 Oelbermann, K., R. Langel, and S. Scheu. 2008. Utilization of prey from the decomposer system by generalist predators of grass land. Oecologia 155: 605617. OzoresHampton, M. 1998. Compost as an alternative weed control method. Hort. Sci. 33: 938940. Parafina, S. 1993. Household relations and agricultural decision making in Haiti. MS thesis, University of Florida, Gainesville, FL. Pascual Villalobos, M. J., A. Lacasa, A. Gonzlez, P. Var, and M. J. Garca. 2006. Effect of flowering plant strips on aphid and syrphid populations in lettuce. Europ. J. of Agron. 24: 182185. Pimentel, D. 1961. Species diversity and insect population outbreaks. Ann. Entomol. Soc. Amer. 54: 7686. Powell, W., M. P. Walton, and M. A. Jervis. 1996. Populations and communities, pp. 225231. In M. A. Jervis and N. Kidd (eds.), Insect natural enemies. Practical approaches to their study and evaluation. Chapman and Hall, London, UK Powers, L. and R.McSorley. 2000. Ecological principles of agriculture. Delmar Thom p son Learning, Albany, NY. Riechert, S. E., and L. Bishop. 1990. Prey control by an assemblage of generalist predators spiders in garden test systems. Ecology 71: 1441 1450. Root, R. B. 1973. Organization of plant ar thropod association in simple and diverse habitats: the fauna of collards ( Brassica oleracea). Ecol. Monogr. 43: 9124. Russel l E. P. 1989. Enemies hypothesis: a review of the effect of the vegetational diversity on predatory insects and parasitoids. Environ. Entomol. 19: 590599. SAS Institute. 2008. The SAS system 9.1 for Windows. SAS Institute, Cary, NC. Schellhorn, N. A., and V. L.Sork. 1997. The impact of weed diversity on insect population dynamics and crop yield in collards, Brassica oleracea (Brassicaceae). Oceologia 111: 233240. Schuster, D. 2006. Pepper weevil: Anthonomus eugenii pp. 5758. In J. L. Gillett, H. N. HansPete rsen, N. C. Leppla, and D. D. Thomas (eds.), Growers IPM guide for Florida tomato and pepper production. UF/IFAS Cooperative Extension, University of Florida, Gainesville, FL.
156 Settle, W. H., H. Ariawan, A. E. Tri, W. Cahyana, A. L. Hakim, D. Hinayana, A. Sri Lestari, and P. Sartano. 1996. Managing tropical rice pests through conservation of generalist natural enemies and alternative prey. Ecology 77: 1975 1988. Sverin, F. 2002. Plant ak pyebwa t d Ayiti. Editions Quitel de Desk Top Adivsory, Por t auPrince, Haiti. Shannon, D. A., J. D. Lea, L. Isaac, and S. Belfort. 2001. SECID/Auburn University PLUS Final Report. USAID/Haiti Economic Growth Office, Washington, DC. Sheehan, W. 1986. Response by specialist and generalist natural enemies to agroecosystem diversification: a selective review. Environ. Entomol. 15: 456 461. Shi, Z H, L D Chen, C F Cai, Z X Li, and G H Liu. 2009. Effects of longterm fertilization and mulch on soil fertility in contour hedgerow systems: A case study on steeplands from the Three Gorges Area, China. Nutr. Cycl. Agroecosys. 84: 3948. Showler, A. T. and S. M. Greenberg. 2003. Effects of weeds on selected arthropod herbivore and natural enemy populations and on cotton growth and yield. Environ. Entomol. 32: 3950. Sivamohan, M. V. K., C. A. Scott, and M. F. Walter. 1993. Vetiver grass for soil and water conservation: prospects and problem, pp. 293309. In D. Pimental, (ed.), World soil erosion and conservation. Cambridge University Press, Cambridge, UK. Smith, H. A. and R. McSorley. 2001. Intercropping and pest management: a review of major concepts. American Entomologist 46: 153161. Smucker, G. R., T. A. White, and M. Bannister. 2002 Land tenure and the adoption of agricultural technology in Haiti. 9th Biennial Conference of the International Association f or the Study of Common Property. International Associati o n for the Study of the Commons, Mexico City, MX. Southwood, T.R.E., and P. A. Henderson. 2000. Ecological Methods. 3rd ed. Blackwell Science Ltd./University Press, Cambridge, UK. Sridhar, M.K.C., G. O. Adeoye, and O. O. AdeOluwa. 2001. Alternate Nitrogen amendments for organic fertilizers. Proceedings of the 2nd International Nitrogen Conference on Science and Policy. Sci. World. 1: 142147. Stoner, K. A. F. J. Ferrandino, M. P. N. Gent, W. H. Elmer, and J. A. Lamondia. 1996. Ef fects of straw mulch, spent mushroom compost, and fumigation on the density of Colorado potato beetles (Coleoptera: Chrysomelidae) in potatoes. J. Econ. Entomol. 89: 12671280.
157 Tahvanainen J. O., and R. B. Root. 1972. The influence of vegetational diversity on the population ecology of a specialized herbivore, Phyllotreta cruciferase (Coleoptera: Chrysomelidae). Oecologia 10: 321246. Thomas, B. P., L. Law, Jr., and D. L. Stankey. 1979. Soil survey of Marion county area, Florida US Department of Agriculture, Soil Conservation Service, Washington D.C. Thomson L. J. and A. A. Hoffman. 2007. Effects of ground cover (straw and compost) on the abundance of natural enemies and soil macro invertebrates in vineyards. Agric. For. Entomol. 9: 173179. Trenbath, B R. 1993. Intercropping for the management of pests and diseases. Field Crops Res. 34: 381405 Trenbath, B. R. 1977. I nteractions among diverse hosts and diverse parasites. Ann. N. Y. Acad. Sci. 287: 124150. Triplehorn, C. A. and N. F. Johnson. 2005. Borror and DeLongs Introduction to the Study of Insects. 7th ed. Thompson Brooks/Cole, Pacific Grove, CA. Tscharntke, T., R., Bommarco, Y. Clough, T. O. Crist, D. Kleijn, T. A. Rand, J. M. Ty lianakis, S. V. Nouhuys, and S. Vid al. 2007. Conservation biological control and enemy diversity on a landscape scale. Biol. Control 43: 294309. (USDA) United States Department of Agriculture. 1989. United States Standards for Grades of Sweet Peppers. Washington, DC http://www.hort.purdue.edu/prod_quality/quality/peperswt.pdf v an Den Bosch, R. 1978. The pesticide conspiracy. Doubleday, Garden City, N Y. van Emden, H. F. 1965. The role of uncultivated land in the biology of crop pests and beneficial insects. Sci. Hort. 17: 121136. Wang, G., M. E. McGiffen Jr., J. D. Ehlers, and E. C. S. Marchi. 2006. Competitive ability of cowpea genotypes with different growth habit. Weed Sci. 54: 775-782. Wiens, T. and C. Sobrado. 1998. Rural poverty in Haiti technical paper 1. In Haiti: The challenges of poverty reduction. V olume II technical papers Report No. 17242 HA. Poverty Reduction and Economic Management Unit and Caribbean Country Management Unit, Latin American and the Caribbean Region. World Bank, Washington, D.C. Wilken, G. 1972. Microclimate management by traditional farmers. Geograph. Rev. 62: 544560.
158 Wilkinson, T. K. and D. A. Landis. 2005. Habitat diversification in biological control: The role of plant resources pp. 305325. In F. L. Wckers, P.C.J. van Rijn, and J. Bruin ( eds. ), Plant provided food and plant carnivore mutualism Cambridge University Press Cambridge, U.K. Wolcott, G. N. 1928. Increase of insect transmitted plant disease and insect damage through weed destruction in tropical agriculture. Ecology 9: 4: 461466. Wyckhuys, K. A. G. and R. J. O'Neil. 2007. Local agroecologi cal knowledge and its relationship to farmers' pest management decision making in rural Honduras. Agric. Hum. Val. 24: 307321.
159 BIOGRAPHICAL SKETCH Heidi HansPetersen graduated with a BS in Biology from North Park University in Chicago, IL, and received her M S in plant pathology and Ph. D in entomology from the University of Florida in May of 2010. She hopes to spend her career working in partnership with low resource farmers and the scientific community to develop relevant and sustainable strategies that increase agricultural production, food security and livelihood systems while conserving the environment. Heidi is a n avid bicycle commuter, canoe enthusiast, and novice beekeeper.