Factors Affecting Windbreak Understory Species Composition: 
An Agroecosystem Interface in Monteverde, Costa Rica

Owen D. Solberg
University of California at Davis
EAP Tropical Biology Program
16 December 1994

Picture of typical pasture land and adjacent windbreaks in the La Cruz area of Monteverde.

Abstract

Windbreaks have been intensively planted throughout Monteverde, Costa Rica. While intended to increase agricultural productivity, they have also become an important component of the non-agricultural ecology of the region. In this study, I examine how physical characteristics and species composition of windbreaks affect the species richness of plants growing beneath them. I find that the percent of light transmitted through a windbreak and the density of trees in a windbreak correlate with understory species richness. I also find that more woody species grow under windbreaks of trees that produce animal-attracting fruits. I speculate that this is caused by greater use of these windbreaks by seed dispersing birds and mammals. In contrast with earlier study and expectations, woody species richness is not greater under native or non-coniferous windbreaks. However, since almost all understory woody species found are native, exotic and coniferous species are probably incapable of reproduction. Therefore, I conclude that in order to provide possible wildlife habitat, to encourage understory woody recruitment, and to help with possible future reforestation, native trees that produce animal-attracting fruits should be planted as windbreaks.

Resumen

Los tapavientos han sido sembrados intensivamente en Monteverde, Costa Rica. Para intentar un aumento de la productividad agricola, ademas ellos tambien seran importantes en la ecologia de la region. En este estudio, examino como las caracteristicas fisicas y la composicion de especies de arboles en los tapavientos afectan las plantas que crecen abajo de los mismos. Encontro que el porcentaje de luz que pasa por los arboles y la densidad de estos se correlacionan con la riqueza de las plantas del sotobosque. Tambien, encontro que mas especies lenozas crecen bajo los tapavientos de arboles que producen frutas y atraen animales. Especulo que esto es causado por aves y mamiferos quienes dispersan las semillas. Concluyo que para asegurar la regeneracion de los tapavientos y talvez la reforestacion en el futuro, los agricultores deben sembrar arboles nativos y los que produscan frutas para los animales.

Introduction

During the last fifteen years, the agroecosystem in the dairy-producing region of Monteverde, Costa Rica, has expanded to include windbreaks. Encouraged by higher pasture productivity and the Monteverde Conservation League's ambitious tree-planting program, agriculturalists planted more than 500,000 trees in long, narrow strips criss-crossing agricultural lands. By protecting pasture and livestock from strong, seasonal wind, windbreaks play a well-documented role in increasing milk production (Ree Sheck, pers. comm. 1994). Thus, this relatively new human-tree mutualism seems assured to last.

Beyond their agricultural purpose, ecologists have become interested in how half a million new trees interact with Monteverde's natural ecosystem. Planted in rows running across pastures, they often connect isolated forest fragments that are so common on these slopes. Therefore, ecologists have speculated that windbreaks might benefit forest creatures by providing temporary refugia and short-distance corridors between forest fragments. DeRosier and Nielsen (1994) found that birds do use windbreaks preferencially over other pasture land. Furthermore, windbreak health might be improved by this interaction with the forest environment: seed-dispersing animals could bring seeds from windbreak trees and nearby forests and promote woody recruitment under windbreaks. In essence, if windbreaks could themselves behave as forest fragments, they could maintain themselves indefinitely. A self-sustaining windbreak would be more valuable than one which senesced and required replanting every 50 to 100 years. Finally, windbreaks could jump-start possible future reforestation efforts. The "nuclear tree" effect describes a mechanism by which trees promote woody regeneration (Janzen 1988). Griggs et al. (1992) demonstrated this effect in the Monteverde area. Bieniek (1992) demonstrated how tree composition of windbreaks can affect woody regeneration. These points indicate the usefulness of windbreaks to conservation biology. But first, we must understand windbreak ecology.

To better understand their ecology, I examined windbreaks and the plants growing under them. I measured various windbreak characteristics and hypothesized that these characteristics influence the species richness and composition of the plants growing under them. In particular, I examined windbreak width, percent light transmittance, windbreak tree density, windbreak tree species richness, percent of windbreak trees with bird or mammal attracting fruit, percent of native windbreak trees, and percent of coniferous windbreak trees. I sought to know if any of these characteristics increased the recruitment of woody seedlings. I assumed that the richness of woody seedlings would indicate a more useful and productive windbreak - to Monteverde wildlife, agriculture, and conservation.

Materials and Methods

I sampled 20 windbreaks from 14 to 27 November 1994 in the Monteverde and La Cruz areas of Costa Rica. I selected windbreaks for study based on the following criteria: grazed pasture fronted the row of trees on both sides; fencing on both sides protected the plants growing under the trees from livestock herbivory; and the ages of the trees were between approximately 5 and 15 years, as determined from farmers. Herbicide treatments, though common in the first few years of a windbreak, are never applied under mature trees (Orlando Varela Duran, pers. comm. 1994). Thus, each windbreak study site had similar treatment histories.

I recorded the species composition of each windbreak study site and later learned the origin and fruiting mechanism of each tree species. I sampled the site 4 times: beginning 10 m from one end of the windbreak, I made a 2x2 m quadrat centered across the width of the trees; I recorded all wild vegetation growing within these 4 square meters; I measured the width of the windbreak at this point; I measured the percent of light transmitted through the trees at this point by taking two readings with a Weston camera photometer, one under the trees and one under open sky; and I repeated this procedure 3 more times, at 10 m intervals. Finally, I counted all the planted trees in the 30 m between my first and last samples. I divided the number of trees by the average width to find the width-wise density of trees.

To characterize windbreak species composition, I classified the tree species: native (to Costa Rica) or exotic; producing animal-attracting fruits ("fruiting") or non-biologically dispersed seeds ("non-fruiting"); and coniferous (Cupressus sp. and Casuarina sp.) or non-coniferous. I then calculated the percentage of trees in each category for each windbreak.

I used a parametric correlation analysis to look for relationships between herbaceous and woody undergrowth diversity and the following 7 independent variables: windbreak width, percent light transmittance, windbreak tree density, windbreak tree species richness, percent of windbreak trees with bird or mammal attracting fruit, percent of native windbreak trees, and percent of coniferous windbreak trees. I also tested for correlations between windbreak width and light transmittance, tree density and light transmittance, and herbaceous and woody species richness.

Results

Table 1 summarizes average characteristics of the samples. Tree density was the most variable physical characteristic, followed by percent light transmittance and width. Of the 17 statistical tests with my data, 5 yielded significant results (Table 2). Light level positively correlated with herbaceous species richness, with the highest variability centered around the approximate mid-value of 65% transmittance (Fig. 1). Light level did not, however, correlate with woody species richness.

Windbreak density positively correlated with herbaceous species richness (Fig. 2). While this may seem to contradict the previous correlation between light level and herbaceous species richness, a separate test showed that density did not correlate with light level. Density also negatively correlated with woody species richness (Fig. 3). Herbaceous and woody species richness responded oppositely to this and to every tested windbreak factor, even those that were statistically insignificant. In fact, herbaceous and woody species richness negatively correlated when tested against each other (Fig. 4).

Windbreak width did not correlate with either herbaceous or woody species richness.

Of the windbreak species characteristics tested, only the percentage of trees with animal-attracting fruits positively correlated with woody diversity (Fig. 5), with many windbreaks of zero fruiting trees showing zero seedlings. The other windbreak species characteristics did not correlate with either herbaceous or woody species richness. Neither the percentage of native trees nor the percentage of coniferous trees correlated with woody species richness, even though there has been speculation as to possible effects of both.

Finally, I observed a general trend in seedling recruitment. Except for 3 less-common species, all woody understory recruitment was of native trees. The exotics found were characterized as pan-tropical weedy trees. Seedlings of truly exotic species, such as Curpessus sp. and Casuarina sp. were never found.

Discussion

Three regression analyses demonstrated significance between windbreak physical characteristics and understory species richness. Light level was positively correlated with herbaceous diversity. This is because at higher light levels, a greater number of the pasture species (the source of many understory species) can grow under windbreaks. Density of windbreak trees was positively correlated with herbaceous diversity and negatively with woody diversity. However, since tree density did not affect light levels, it is unclear how density affected understory species richness. It may be that the less dense windbreaks were able to grow taller and produce more litter, which might shift the competitive advantage from herbaceous to woody species.

The negative correlation between herbaceous and woody species richness is probably due to different optimum conditions and competition between these two groups of plants. I surmise that the herbaceous species, being far more numerous and widespread, affect the diversity of the woody species. In other words, the woody species are outcompeted when the herbaceous species become very widespread.

I believe I found no correlation between windbreak width and understory species richness primarily because the widths were relatively uniform. Any windbreaks within my sampled range of 2.2 to 6.5 m are probably equally regarded by weedy herbs that can disperse from tens to hundreds of meters.

Windbreaks can undoubtedly affect the understory vegetation in many ways. But like other investigators, I believed the most interesting effects would be due to windbreak species composition. The percentage of trees with animal-dispersed seeds in a windbreak was positively correlated with understory woody species richness. Other studies have demonstrated the role of trees in attracting seed dispersers (DeRosier 1994) and promoting woody regeneration (Janzen 1988; Griggs et al. 1992). Therefore, I suggest that more woody species were found beneath these trees because birds or mammals use these windbreaks for habitat or food. I assume that these seed dispersers brought seeds with them from other habitats, since the woody species found were usually different from the windbreak species. No correlation linked these animal-dispersed windbreak trees with herbaceous species richness, possibly because the herbaceous species do not rely on the same vertebrate dispersers that are attracted to the windbreak trees.

No correlation was found when the percent of native windbreak species was compared with understory diversity. Although recent interest in the ecological benefits of planting native trees instead of exotics, increasing the diversity of understory seedlings does not appear to be one of them. Perhaps this is not suprising, since ecologically, an exotic species may be "naturalized" by relationships with native species. In effect, an animal may not care where a tree came from, only what food or habitat it can provide.

Likewise, no correlation was found between the percent of coniferous windbreak species and understory diversity. Given that this relationship has been previously supported (Bieniek 1992), this result is curious. The discrepancy may be due to variations in sampling and analysis. The Bieniek study counted only native understory species and included some monospecific and younger windbreaks.

Finally, almost no understory seedlings were exotics (even when the windbreaks were composed of exotics), and none were coniferous. Janzen (1986) describes species he calls the "living dead." Throughout the tropics, there are species still living that have lost their ability to reproduce, usually due to human-caused habitat alterations. Although the concept may be intended for species like figs and hawks, he notes that it may be applied to agricultural species, too. In this study, the majority of exotic, planted trees were probably living dead species because they showed no seedling recruitment. These are trees which, though they may be living and working for the farmers today, will not pass on future generations of windbreak trees.

With this small addition to our understanding of windbreaks, I make several recommendations. To promote seedling species richness beneath windbreaks, I suggest that windbreaks be planted at medium to low density levels. I suggest that the species used for windbreaks should be trees that make animal-attracting fruits and trees that are native. These fruiting trees will attract birds and mammals by providing food and refugia. Also, if these trees are planted around hotels, a higher abundance of foraging birds may please tourists. I suggest native trees because only they have demonstrated a reproductive future. Together, these recommendations may help to improve wildlife habitat, windbreak longevity, and possible future reforestation efforts.

Aknowledgements

I thank Bill Haber for graciously serving the community in the dissimination of a vast mental botanical database, Alan Masters, Federico Chinchilla, and Frank Joyce for sage guidance, Chad Steiner for statistics and music, Orlando Varela Duran and Ree Sheck of the Monteverde Conservation League for invaluable information, Nuria Vargas for food and food for thought during the study period, the staff of the Estacion Biologica for amiable excellence, Colin Lee and Esther Selk for needed feedback on this paper, Sylvia Fallon for enjoyable lizard-hunting expeditions, and Jon and Sarah and Newt.

Literature Cited

Bieniek, M. 1992. A comparative study of regeneration in conifer and non-conifer windbreaks. UCEAP #92-55.

DeRosier, D. and K.B. Nielsen 1994. Biological assessment of avian usage of agricultural windbreaks in the Monteverde zone of Costa Rica. Preliminary statement of findings. Unpublished manuscript.

Griggs, J., J. Hansen, J. Lillien, M. Stauber, and D. Zeilinger 1992. Patterns of seed dispersal and forest regeneration along the forest edge, around nuclear trees, and in open pastures. UCEAP #92-5.

Janzen, D.H. 1986. The future of tropical ecology. Ann. Rev. Ecol. Syst. 17:305-24.

-----------. 1988. Management of habitat fragments in a dry tropical forest: growth. Ann. Miss. Bot. Gard. 75:105-116.