I found Bagnalliella mojave Hood (Thysanoptera: Phlaeothripidae) restricted to small, isolated Yucca brevifolia Engelmann (Liliales: Agavaceae) that likely compete less for soil moisture and have high leaf water-contents.
Bagnalliella mojave Hood (Thysanoptera: Phlaeothripidae) (Hood 1927) is a wing-dimorphic thrips found only on Yucca brevifolia Engelmann (Liliales: Agavaceae) (Cott 1956). Yucca brevifolia, or Joshua tree, is a large, arborescent monocot endemic to the Mojave Desert in the southwestern U.S. The plant produces 1–3 branching, trunk-like stems that terminate in clusters of rigid, semi-succulent, pointed leaves (Webber 1953). Bagnalliella mojave lives between young leaves at the center of leaf clusters.
Joshua trees appear to compete for soil moisture, because plant size increases with increasing spacing from other Yucca plants (Yeaton et al. 1985). Availability of soil moisture likely affects the water content, growth, and size of plants. Bagnalliella mojave may require plants with high water-contents, because it is more abundant on younger Y. brevifolia and on more-succulent leaf-clusters on older plants (Cott 1956). I evaluated the dependence of B. mojave on host-plant water-content by examining its presence or absence on different sizes of Y. brevifolia growing together with Yucca schidigera Roezl ex Ortgies.
The study was conducted at the southern end of the Eldorado Mountains, 5.4 km northeast of Searchlight, Clark County, southern Nevada. The area (35°30′N, 114°53′W; elevation 1,140 m asl) contains outcroppings of weathered granite interlaced with washes. Yucca brevifolia at the study area have the short growth-form of Yucca brevifolia var. jaegeriana McKelvey (Fig. 1a). Rainfall at Searchlight averages 196 mm yearly, mostly during Dec-Mar and Jul-Sep (DRI 2014), and totaled 102 mm in 2011, 184 mm in 2012, and 177 mm in 2013 (CCRFCD 2014).
I searched for B. mojave during 2011–2013 by examining leaf clusters on differently-sized Y. brevifolia and recorded the locations of the 12 plants found to be inhabited. The rare aggregations of the thrips were recognized by the relatively large, apparentlyblack adults and frequently-present, red immatures. Wing morphs of adults, some treated with NaOH, were mounted on slides and identified as B. mojave following Cott (1956). I compared mounted females with those at the Entomology Research Museum, University of California, Riverside, and deposited vouchers (nos. 417188–417192). I photographed life stages under incident light and superimposed images at different focal planes with CombineZP (Hadley 2013). Second-instar larvae and brachypterous (Fig. 1f) and macropterous adults were collected from a single leaf-cluster during Jul 2011. First- and second-instar larvae (Fig. 1c,d), one propupa (Fig. 1e), and brachypterous adults were collected from a second leaf-cluster during Oct 2013. Both larval instars and brachypterous adults were collected from a third leaf-cluster during Jan 2014. Leaf clusters on Y. brevifolia likely support multiple generations of B. mojave, because leaf clusters on the first plant (Fig. 1a) found to contain B. mojave during Jul 2011 still contained thrips during Dec 2013. The single propupa (third instar) that was found in an aggregation of larvae and brachypterous adults suggests B. mojave pupates and completes its life cycle within the same leaf-cluster.
Sizes and spacing of Y. brevifolia plants inhabited or not inhabited by B. mojave were measured during 2013–2014. Each recorded plant was relocated, and the continued presence of the thrips on at least 1 leaf-cluster was verified with a hand lens. Plant size was measured by counting the number of leaf clusters. Plant spacing was measured as the distance from the inhabited plant to the nearest Y. schidigera and Y. brevifolia in 4 quarters (compass bearings N-E, E-S, S-W, W-N; point-centered quarter method, Greig-Smith 1964). The four, nearest Y. brevifolia were sampled for B. mojave by examining all leaf clusters that contained young leaves at or below eye-level (34 plants). Plants with all leaf-clusters above eye-level (14 plants) were not sampled. At each of the nearest Y. brevifolia sampled for thrips, I similarly counted the number of leaf clusters and measured the distances to the nearest Yucca plants. Bagnalliella mojave were found on 2 of the nearest Y. brevifolia, producing 14 plants inhabited by B. mojave and 32 plants not inhabited by B. mojave.
Plant size was regressed (Systat version 10.2, Chicago, Illinois) against 4 measurements of plant spacing: (1) the mean distance (across the 4 quarters) to Y. brevifolia, (2) the mean distance to Y. schidigera, (3) the mean distance to the nearest Y. brevifolia or Y. schidigera in each quarter, and (4) the mean distance to Y. brevifolia and Y. schidigera within and across quarters. Numbers of leaf clusters were transformed log (N), and distances between plants were transformed log (m), to normalize residuals. Plant size was most related, positively, to mean distance to the nearest Yucca of either species in each quarter (F = 11.6; df = 1,44; P < 0.001; R2 = 0.21; Fig. 2).
Mean distance to the nearest Yucca did not differ (t = 0.73; df = 44; P = 0.47) between plants inhabited (6.1 m, back-transformed) or not inhabited (6.6 m) by B. mojave (Fig. 2). Adding thrips presence or absence as an indicator variable to the regression of plant size against plant spacing significantly decreased the error variance (F = 70.3; df = 1,43; P < 0.001; partial R2 = 0.53; Fig. 2). The 14 inhabited plants supported 2–16 leaf clusters (back-transformed mean = 5.8 clusters), and the 32 uninhabited plants supported 8–146 leaf clusters (33 clusters). Eleven plants, 6 with thrips and 5 without thrips, overlapped with 8–16 leaf clusters.
Yucca brevifolia plants inhabited by B. mojave are smaller than expected based on their spacing from other Yucca plants and likely compete less for soil moisture. This agrees with Cott's (1956) observation that the thrips occurs mostly on younger plants. Absence of B. Mojave in more-succulent leaf-clusters on older, larger plants, as described by Cott (1956), may have been due to lower than average rainfall. Bagnalliella mojave appears to be a rare phytophagous thrips mostly limited to small, isolated Joshua trees with high leaf water-contents.