NATURE|Vol 436|11 August 2005 NEWS & VIEWS 783 depend on solar activity. That speed exceeds by about a factor of ten the speed of sound and the speed of the most common magnetic waves, known as Alfvén waves. (Alfvén waves are highly correlated fluctuations in both the fluid-velocity and magnetic fields.) As the supersonic, super-Alfvénic solar wind encounters Earth’s magnetic field, a bow shockwave is produced at about 10–15 Earth radii in front of Earth (Fig. 1). Behind the bow shock, the hot solar-wind plasma can flow down towards the ionosphere through the dayside cusp. This cusp forms the bound- ary between magnetic field lines that are closed on the dayside (the side of Earth exposed to the Sun) and magnetic field lines that are open and have been swept back into the lobes of the nightside magnetosphere. The flow of plasma down the funnel-like cusp has been conjectured either to excite tur- bulence locally or to amplify the turbulence carried by the shocked solar-wind plasma. Savin et al. 2 noted that the flow down the cusp should generate vortices. Nykyri et al. 3 found, using Cluster magnetometer data from March 2001, evidence that the cusp contained mag- netic turbulence. Cluster’s four spacecraft orbit such that, at the point where they are farthest from Earth — at apogee — their positions form a regular tetrahedron (Fig. 1). This formation is ideal for distinguishing between unchanging spatial features and features that evolve with time. In the spring of 2002, at the time of Sundkvist and colleagues’ measurements 1 , the separation of the spacecraft at apogee was about 100 kilo- metres (compared with around 500 kilometres for earlier Cluster measurements 3 ), allowing the resolution of smaller spatial features. When the spacecraft entered the cusp, they were still relatively close together, but no longer traced out the points of a regular tetra- hedron (see Fig. 3e on page 827). Turbulence is often described as a process in which large-scale eddies cascade down to smaller-scale eddies until a scale is reached at which dissipation sets in. In magnetized plasmas, because of the large variety of possi- ble small-scale wave modes, it is not clear how that cascade progresses to the dissipation range. Understanding these processes would enable us to determine how energy flows in a turbulent magnetofluid from large scales to the smaller, kinetic scale — and thus heats Earth’s ambient plasma. Furthermore, where vortices form, materials in initially separate regions of space become mixed, which transfers energy, momentum and mater- ial from one region of the magnetosphere to another. In the solar wind itself, measurements from single spacecraft 4,5 indicate that as the scales of the turbulent fluctuations approach the dissi- pation scale, the kinetic Alfvén wave becomes the predominant wave mode. The distinguish- ing characteristic of such waves is that they have a small electric field that is parallel to the direction of the local magnetic field. Sundkvist and colleagues 1 have now analysed the tempo- ral evolution of the magnetic field in the day- side cusp region. There they show that kinetic Alfvén waves interact nonlinearly with so- called drift waves caused by gradients in plasma density and magnetic fields, and pro- vide evidence that this interaction produces distinctive small-scale turbulent features known as drift–kinetic Alfvén vortices. Mea- surements of the velocity shears across the plasma flow direction made by Cluster’s ther- mal plasma instrument indicate amplitudes that exceed those required for vortex produc- tion. The authors’ interpretation is further bolstered by a vortex model constructed by the authors that accurately reproduces the observed behaviour of the magnetic field. The Cluster observations are the first mea- surements in space to indicate that small-scale vortices are formed as eddies reach the dissi- pation scale. At the time of the observations, Cluster 1 and 2 were aligned with the plasma flow. Data from those two spacecraft indicated that the observed structures were quasi- stationary. The other two spacecraft were not aligned with the flow and could be used to deduce that the transverse radial scales of the structures were a few proton gyroradii (the radius of the circle described by a proton mov- ing across a background magnetic field — in this region of the cusp, about 25 kilometres). The small spatial scales involved make mea- surements of a turbulent cascade’s dissipation — and of the transfer of the energy contained in magnetic fields and particle motion into the heating of the ambient plasma — difficult in both terrestrial laboratories and in space. Cluster observations indicate the existence of a turbulent vortical boundary layer that enhances the transfer of momentum and energy from the solar wind to the magneto- sphere. The four-spacecraft Cluster, with its ability to distinguish between spatial and tem- poral effects, has opened a new window on the study of turbulence, both in the magneto- sphere and in the solar wind. In the near future, missions such as Magnetospheric Multiscale (MMS), with its even smaller spacecraft sepa- ration and higher time-resolution for plasma measurements, will further enhance our understanding of the generation and dissipa- tion of magnetofluid turbulence. ■ Melvyn L. Goldstein is at the NASA Goddard Space Flight Center, Code 612.2, Greenbelt, Maryland 20771, USA. e-mail: melvyn.l.goldstein@nasa.gov 1. Sundkvist, D. et al. Nature 436, 825–828 (2005). 2. Savin, S. P. et al. JETP Lett. 74, 547–551 (2001). 3. Nykyri, K. et al. Ann. Geophys. 22, 2413–2429 (2004). 4. Leamon, R. J. et al. Astrophys. J. 537, 1054–1062 (2000). 5. Bale, S. et al. Phys. Rev. Lett. 94, 215002 (2005). PLANT BIOLOGY Engineered male sterility Muhammad Sarwar Khan The phenomenon of ‘cytoplasmic male sterility’ in plants has long been exploited to enhance the productivity of certain crops. An innovative genetic-engineering system promises to widen applicability of the approach. Among the main items on the wish-list of plant breeders are these. First, the ability to artificially suppress pollination and so prevent a plant’s self-fertilization, thereby encouraging cross-pollination and higher-yielding seed through an effect known as ‘hybrid vigour’. Second, the ability to genetically engineer such suppression of male fertility into elements in the cytoplasm, rather than the nucleus — the result is transmission of desirable characteris- tics through genes in the female line, and those genes cannot ‘escape’ uncontrollably via pollen. Third, the ability to selectively restore male fertility. Although farmers want high- yielding hybrid seeds, for certain crops that seed has to produce fertile plants. These aims can be achieved by exploiting a phenomenon known as ‘cytoplasmic male sterility’. As they describe in Plant Physiology, Ruiz and Daniell 1 have demonstrated a promi- sing new way of achieving all three goals. Their approach was tested in tobacco plants. It involves inserting a gene (phaA) from the bac- terium Acinetobacter into plant chloroplasts, with — upstream of that gene — a ‘promoter’, psbA, and other regulatory elements. The phaA gene encodes an enzyme, ȋ-ketothiolase. The authors show that in their transgenic plants the enzyme accumu- lates in leaves and anthers, the pollen- producing structures, and alters the course of synthesis of fatty acids. A starting point in fatty-acid synthesis is acetyl-CoA, which under normal circumstances is converted to malonyl-CoA. However, ȋ-ketothiolase over- rides the usual enzyme involved, acetyl-CoA carboxylase, to produce acetoacetyl-CoA instead (Fig. 1a, b, overleaf). Correct lipid metabolism is essential to the normal devel- opment of pollen, not least the pollen wall 2,3 . Ruiz and Daniell 1 found that ȋ-ketothiolase accelerates anther development and, among other consequences, causes the pollen grains to collapse — and thereby results in male Nature Publishing Group ©2005