Effect of honey mesquite size and distribution on desert mistletoe parasitism

Angelique Leonard1, Monica Rivas2, Thomas Savoie3, Olivia Zanzonico4

1University of California, Merced; 2University of California, Los Angeles; 3University of California, Santa Cruz; 4University of California, Davis

Parasitic plants must rely on their hosts to survive. However, the distribution and survival of parasitic plants often depends on other organisms in addition to the host. Desert mistletoe (Phoradendron californicum) has a unique interdependent relationship with its preferred host, honey mesquite (Prosopis glandulosa), and its primary disperser, Phainopepla (Phainopepla nitens). We sought to study this relationship by testing the effects of honey mesquite size and distribution on the likelihood of infection by mistletoe. At four different sites in and around Anza Borrego State Park, we measured the size of honey mesquite thickets and distance between thickets. We found that there was no effect of distance between host and neighbors on the presence of mistletoe. However, the host height, volume of neighboring mesquite thickets, and height of neighboring thickets had a positive relationship with mistletoe presence. In addition, there was a noticeable effect of sampling site on the presence of mistletoe. The two sites near a source of water had a higher percentage of infected hosts than the two without a water source. Implications of our findings may include conservation of Phainopepla in the future by protecting desert environments where honey mesquite is found by water sources, thereby increasing presence of mistletoe.

Keywords: Phainopepla nitens, Phoradendron californicum, Prosopis glandulosa, plant parasite, spatial distributions

INTRODUCTION 1992). Plant parasites have negative effects on their hosts, ranging from a decrease in Host-parasite interactions can give way to size, limited water, reduced nutrients, and an evolutionary arms race that establishes a even death but often depend on biotic or complex network of unique vectors, hosts, abiotic factors in order to spread (Yule and and parasite dispersal strategies. Bronstein 2018, Candia et al. 2014, Understanding the spatial patterns that Mathiasen et al. 2008). lead to the spread of a parasite can provide Mistletoes (Viscaceae) are a highly diverse insight into the dynamics which drive these group of aerial hemi-parasites that rely on interdependent relationships (Kareiva animal vectors to pollinate and disperse 1994). Patterns of the spread of a disease them (Aukema 2003). Hemi-parasites are are dependent on the spatial distribution of capable of photosynthesis but need a host both host and parasite, as well as vector for water and nutrients (Calder 1983). prevalence and preference (Real et al. Desert mistletoe (Phoradendron californicum)

CEC Research | https://doi.org/10.21973/N39M1P Fall 2019 1/7

depends upon pollinators, mainly insects Phainopepla can more easily see their and birds, to exchange genetic material, territorial displays, therefore leading to and dispersers which determine the higher rates of mistletoe infection on taller location of the establishment of offspring mesquite (Aukema 2004). However, little is on hosts (Yule and Bronstein 2018, known about how other spatial factors of Mathiasen et al. 2008). In the desert, mesquite plants may contribute to the mistletoe maintains a mutualistic distribution of desert mistletoe. relationship with its vectors, as it is one of In this study, we investigated the effects the only available resources during its fall- of honey mesquite size and density on winter fruiting period (Aukema and desert mistletoe presence across the Martinez 2002). Perhaps its most landscape. We hypothesized that honey specialized relationship is with the mesquite thickets that were both larger and Phainopepla (Phainopepla nitens). growing closer together would be more Phainopeplas are drawn to the Sonoran likely to be infected by mistletoe, due to Desert in the winter to breed and gorge Phainopepla preference to exert less energy themselves on mistletoe berries (Aukema flying between sparsely growing hosts. We and Martinez 2002). A sticky viscin coats the also predicted that infection rates would be berries, allowing them to adhere to higher in larger thickets due to increased branches after being deposited by area available for mistletoe infection. defecation, regurgitation, or bill wiping (Reid et al. 1995). Phainopepla have a METHODS digestive system specialized for mistletoe berries, and in the desert will defend 2.1 Site Description territories of host plants infected with Sampling for this study was conducted at mistletoe (Aukema 2004). four sites in and around Anza Borrego State Desert mistletoe has several hosts in the Park in early November 2019. Anza Borrego Fabaceae family in the Sonoran Desert, but State Park is in the Sonoran Desert in is most successful on mesquite, receiving southeastern California, at the eastern edge more pollen, growing larger, and with more of the Colorado Desert. Field samples were berries than on trees such as acacia (Yule taken at four sites adjacent to San Felipe and Bronstein 2018). Yet mistletoe, in Road (San Felipe), Borrego Springs Water contrast, can have a number of detrimental Treatment Ponds (Water Treatment Ponds), effects on mesquite, killing branches, Yaqui Wells Road (Yaqui Wells), and the reducing growth and fecundity, and Great Southern Overland Stage Route sometimes killing the host (Hawksworth (Stage Route). Two of the sites, Yaqui Wells 1983). Honey mesquite (Prosopis and Water Treatment Plant, were located glandulosa) is a prime host for mistletoe near a water source, a small seep and water due to its sturdy branches and deep root treatment settling ponds, respectively. system which optimizes water uptake in the Vegetation varied slightly among sites but dry desert (Aukema and Martinez 2002). was primarily dominated by honey Previous study has shown that mesquite, catclaw acacia (Senegalia Phainopeplas prefer to defend territories on greggii), and creosote bush (Larrea taller mesquite trees because other

CEC Research | https://doi.org/10.21973/N39M1P Fall 2019 2/7

tridentata). Desert willow (Chilopsis Longest side and height measurements linearis), alternatively, was only found at were also recorded for the two nearest Yaqui Wells. Sampling was performed neighbors. Neighbors were only recorded if approximately 1500 meters south of Hwy they were also honey mesquite. 78 on the Great Southern Overland Stage Route and along a 1500-meter stretch of 2.3 Bird Count San Felipe Road located 428 meters north of Hwy 78. Twelve host thickets and their Estimates for the population of two closest neighbors were sampled at Phainopepla nitens at each site were each site. However, some host thickets did performed using a point count. At each of not have a neighbor within 50 m and were the four sites, point counts were done two considered isolated. As a result, a total of hours after sunrise for a five-minute period. 48 host thickets and 83 neighbors were All Phainopepla that were seen or heard sampled overall. were included in the count.

2.2 Mesquite Selection and Measurement 2.4 Statistical Tests

For each sample, a host patch of honey Statistical tests were performed using mesquite was randomly selected at a JMP 14 software. We utilized a multiple minimum distance of 50 m from other host logistic regression to test the effect of site, patches. Patch height, length of longest host height, average neighbor height, size, and mistletoe presence were average neighbor volume, and average measured. Host honey mesquite plants neighbor distance on the presence of were both individuals and patches of many mistletoe on the host plant. Average individuals. The honey mesquite sampled nearest neighbor data was log transformed was categorized as a single patch so long as to make it normally distributed. We the individuals were no further than one executed a multiple linear regression to test meter apart. Mistletoe infection was the effect of the same factors on mistletoe categorically recorded as either present or density on a host plant. The linear regression absent for each host patch. Mistletoe plants excluded samples with no mistletoe present. could easily be seen and identified on the RESULTS host thicket. Mistletoe was not recorded for neighboring plants. 3.1 Site Dependent Effects The longest side was measured using a rangefinder and used as a proxy for area We found a significant difference in covered by the patch, rather than mistletoe presence among our four sample measuring both sides. When the longest sites (Figure 1, Table 1). The Water Treatment side was not measurable using the Ponds had the highest proportion of host rangefinder due to dense vegetation, we patches infected by mistletoe (83%), used GPS markers to find distance on followed by Yaqui Wells (67%), San Felipe Google Earth. Height measurements of the (42%) and Stage Route (8%). Among host thicket were estimated using a two-meter mesquite hosts with mistletoe presence, site stick and rounded to the nearest half meter. had no effect on mistletoe density (Table 2).

CEC Research | https://doi.org/10.21973/N39M1P Fall 2019 3/7

3.2 Effects of Honey Mesquite Spatial Distribution

Average neighbor distance had no effect on mistletoe presence (Table 1). However, all other spatial factors we tested influenced the probability that a host mesquite thicket would be infected (Table 1). Hosts that were taller were more likely

to have mistletoe present. Hosts with taller Figure 1. Effect of sample site on desert mistletoe neighbors and neighbors of a larger volume presence in honey mesquite. Honey mesquite were also more likely to have mistletoe thickets were sampled at four different sites in Anza Borrego State Park and each thicket sampled was present. Host height, average nearest marked as either having desert mistletoe present or neighbor height, average nearest neighbor absent. Percentages of honey mesquite thickets with volume, and average neighbor distance had mistletoe present were then generated from our no effect on mistletoe density among hosts samples at each site. There was a considerable already infected (Table 2). difference in mistletoe infection between sites (Table 1). 3.3 Birds

Yaqui Wells had the highest number of Phainopepla present at thirteen, followed by the Water Treatment Ponds with eleven, San Felipe with three, and Stage Route with

one. The sample size was too small to Table 1. Effect of site and honey mesquite spatial statistically test any patterns between birds variables on presence of mistletoe on host. Note: and mistletoe presence or density. These values were generated using a multiple logistic regression. Average neighbor distance was log transformed to make it normal (N = 48, 12 per DISCUSSION site, 4 sites). We predicted that greater density of honey mesquite would lead to higher occurrence of mistletoe on honey mesquite. We found no effect of the distance between mesquite thickets on the presence of mistletoe but found that larger neighboring thickets made a host more likely to be Table 2. Effect of site and honey mesquite spatial variables on host mistletoe density. Note: These infected. In addition, larger host plants values were generated using a multiple linear were more likely to be infected. The size of regression. Average neighbor distance was log mesquite thickets varied by site, which we transformed to make it normal (N = 48, 12 per site, have attributed to variability in water 4 sites). proximity. Yaqui Well has a natural water seep that has contributed to it historically providing a water source for desert flora

CEC Research | https://doi.org/10.21973/N39M1P Fall 2019 4/7

and fauna, while the Water Treatment mesquite distance on the likelihood of center offers desert life proximity to water infection. Because Phainopepla can fly by way of human-made water treatment varying distances to and from thickets, the ponds. These two sites with greater distance between thickets may not play a availability of water had much bigger role in the distribution of mistletoe. In our neighboring thickets, and high occurrence study, we found no effect of spatial of mistletoe. This can be explained by the variables measured on the density and/or inherent need for greater water availability abundance of mistletoe within hosts. in order for mistletoe to parasitize the host. However, another study found Mistletoe establishes itself on host plants Phainopeplas drive a positive feedback loop and uptakes water (Knutson 1983). in the abundance of mistletoe in hosts due Therefore, lack of access to water would to their territorial nature in desert habitats more easily kill the parasitized host, (Aukema 2004). Phainopepla tend to favor preventing further proliferation of trees which are already parasitized and mistletoe, as was observed in the much deposit seeds in areas with high mistletoe lower mistletoe occurrence at Stage Road density (Aukema 2004). This may contribute and San Felipe Road. to the pattern of presence and absence of A previous study found that mistletoe mistletoe which we observed. A previous prevalence was more variable in lower study found that Phainopepla tend to elevations and at lower tree densities defend territories with taller mesquite trees (Aukema 2004). However, we found no (Aukema 2004). We found that the sites effect of spatial distribution of honey which had mesquite thickets with greater mesquite on the prevalence of mistletoe, volume also had higher counts of and instead found that variation in Phainopepla. mistletoe abundance was due to the size of Mesquite thicket sizes have increased neighboring thickets independent of since the late nineteenth century and is distance apart. Spatial distribution of honey hypothesized to be due to fire suppression mesquite differed at each site, which can be and dispersal of seeds by livestock (Ansley attributed to differences in local 1997). In this study, we demonstrated that topography. Yaqui Well is located in a this may also be linked to human-made shallow canyon, while the San Felipe site water sources. Future studies may was located at the base of a hilly range. We investigate the effects of these water found that mesquite in these two sites sources on densities within mesquite tended to grow furthest apart from each stands, and whether this may contribute to other, while the mesquite in the flatter, greater abundances of mistletoe. In the more open areas of Stage Road and the past, breeding failures of Phainopepla have Water Treatment site tended to be closer been linked to the absence of mistletoe together. berries due to drought (Chu 1999). Besides host compatibility and Mitigating the effects of extreme drought distribution, Phainopepla exert an with artificial water sources may aid in inexorable influence on the dispersal of management efforts to preserve this mistletoe, which can serve as an characteristic bird. In addition, if proximity explanation for the lack of effect of to water increases abundance of mistletoe,

CEC Research | https://doi.org/10.21973/N39M1P Fall 2019 5/7

conservation efforts may target regions ACKNOWLEDGEMENTS with water sources to preserve other mutualistic pollinators and dispersers of This work was performed at the mistletoe. Further study may investigate University of California’s Steele/Burnand the details of Phainopepla territorial Anza-Borrego Desert Research Center, doi: boundaries and investigate the distribution 10.21973/N3Q94F. A big thank you to our and abundance of mistletoe within these instructors Tim Miller, Krikor Andonian, and boundaries. Additionally, we suggest Reina Heinz for the guidance, assistance, investigating the threshold for mistletoe and encouragement given during this abundance and its viability on a mesquite course. To CEC Fall 2019 cohort, we thank host, and whether there is an “optimal” you for the unlimited laughs, unwavering abundance of mistletoe infection of the spirit, and constant support throughout this host for reproductive success of the process. parasite. Understanding the spatial distribution of REFERENCES mistletoe and its hosts may give insight into Aukema, J. E. 2003. Vectors, viscin, and viscaceae: the structure and function of the mistletoes as parasites, mutualists, and resources. ecosystems they influence. Plant parasites Frontiers in Ecology and the Environment 1:212. such as mistletoe are exceptionally diverse but remain understudied, but because they Aukema, J. E. 2004. Distribution and dispersal of desert mistletoe is scale dependent, hierarchically are long lived, sometimes living nearly as nested. Ecography 27:137–144. long as their hosts, they can provide insight into the coevolution of parasites, hosts, and Aukema, J. E. Martinez, C. 2002. Where does a fruit- vectors. Furthermore, parasitism can serve eating bird deposit mistletoe seeds? Seed as a predictor for increases in diversification deposition patterns and an experiment. Ecology 83:3489–3496. rates due to reproductive isolation following colonization of a new host species Candia, A. Medel, R. 2014. Indirect positive effects of (De Veinne et al. 2013). This has cascading a parasitic plant on host pollination and seed effects on animal vectors which interact dispersal. Oikos 123:1371–1376. with these parasite-host relationships. Plant Chu, M. C. 1999. Ecology and breeding of parasites, and parasites in general, may be Phainopeplas (Phainopepla nitens) in the desert regarded as pests and detrimental to the and coastal woodlands of southern California. ecosystems that they inhabit because of Ph.D. thesis, Univ. of California at Berkeley. their negative effects on hosts, but they serve as important resources for the De Vienne DM, Refrégier G, López-Villavicencio M, Tellier A, Hood ME, Giraud T. 2013. Cospeciation organisms that have evolved to depend on vs host-shift speciation: methods for testing, them, and can play a key role in shaping the evidence from natural associations and relation to ecological relationships around them. coevolution. New Phytol. 198:347–385

Hawksworth FG. 1983. Mistletoes as forest parasites. In: Calder Mand Bernhardt P (Eds). The biology of mistletoes. San Diego, CA: Academic Press. p 317– 33.

CEC Research | https://doi.org/10.21973/N39M1P Fall 2019 6/7

Kareiva, P. 1994. Space: the final frontier for ecology: populations, communities and ecological theory. Ecology 75:1. ecosystems. Chapman and Hall, pp. 492–508.

Knutson, D. M. 1983. Physiology of mistletoe Reid N, Smith NM, and Yan Z. 1995. Ecology and parasitism and disease responses in the host. In: population biology of mistletoes. In: Lowman MD The Biology of Mistletoes. M. Calder and P. and Nadkarni NM (eds). Forest canopies. San Bernhardt, eds. Pp. 295–316. Diego, CA: Academic Press. p 285–310.

Mathiasen, R. L., Nickrent, D.L., Shaw, D., Watson, Shearer, G., D.H. Kohn, R.A. Virginia, B.A. Bryan, J.L. D.M., 2008. Mistletoes: pathology, systematics, Skeeters, E.T. Nilsen, M.R. Sharifi and P.W. Rundel. ecology, and management. Plant Disease 92:988– 1983. Estimates of N2-fixation from variation in 1006. the natural abundance of 15N in Sonoran Desert ecosystems. Oecologia 56:365–373. Real, L., Marschall, E. and Roche, B. 1992. Individual behavior and pollination ecology: implications for Yule, K. Bronstein, J. 2018. Reproductive ecology of a the spread of sexually transmitted plant diseases. parasitic plant differs by host species: vector In: DeAngelis, D. L. and Gross, L. J. (eds), interactions and the maintenance of host races. Individual-based models and approaches in Oceologica 186:471–482.

CEC Research | https://doi.org/10.21973/N39M1P Fall 2019 7/7